Metagora Training Materials

Sustainable Energy Development Office

public policy

Data collected from a subset of a population cannot be considered representative of that population unless the subset of the population from which the data were gathered is a random sample of that population. For example, sometimes an opinion poll is given to the readers of particular magazine. The statistics developed from that opinion poll represent only the viewpoints of the readers of the magazine who chose to take the poll, not the entire population that has access to the magazine, or even the entire pool of readers of that magazine.
When a statistic is created from collected data (for example, the average salary of adult males in the European Union), a confidence interval is usually included in order to provide the reader with an understanding of the margin of error for that statistic. The confidence interval is usually found by multiplying the calculated standard error for the statistic by a constant determined by the level of confidence desired (e.g., for a 95% confidence interval that constant is 1.96). The statistical assumption made when such a confidence interval is calculated is that the data used to create the statistic have a relatively symmetric distribution. The more skew the distribution of the underlying data has, the less valid the statistical assumption of symmetry becomes and the less valid the confidence interval is.

The important lesson is that if the statistical assumptions underlying an analysis are wrong, that analysis may not be valid. Common causes of violations of statistical assumptions include bias, methodology errors, lack of random assignment in the design of an experiment, and lack of random sampling for a survey.

Further information on assumptions can be obtained as follows:

1. From www.thefreedictionary.com/assumption (19 December 2006).

2. Sources for this encyclopedia entry include Wikipedia (accessed 19 December 2006) [disclaimer]; and Helberg, Clay, "Pitfalls of data analysis," from Practical Assessment, Research & Evaluation (1996) and obtained from http://PAREonline.net/getvn.asp?v=5&n=5 (accessed 19 December 2006).

Back-Translation, Translation
See 'Translation, Back-Translation' Below

Barplot
Barplot

Example: Number of human rights violations per type of non-physical ill-treatment

The survey results correspond to 2,300,000 contacts with police forces experienced by 1,600,000 persons last year (México City D.F., 2005).

In statistics, a barplot is a generic graph that uses bars (rectangles) to show the distribution of data representing a population characteristic or a set of population characteristics. There are many types of bar diagrams, but a main distinction between them is the level of measurement of the population characterstic being displayed. For example, histograms are a type of barplot.

In the example above, the data are categorical: mainly, the number of human rights violations per type of ill-treatment. For displaying categorical data such as these, the bars can be vertical or horizontal, and are usually non-adjacent.

The most widely used bar plots are: one bar for each of the categories of some population characteristic (simple chart, like the one above); the representation of several categories within the same bar (stacked or composed, as given in the first graphic below); and several bars presented together (clustered) to represent groups of population characteristics to be compared (as in the second graphic below).

Canada-Ontario Agreement

Stacked barplots can also be given in percentages. In those cases, all bars have the same length, allowing easier comparison of the proportions of different characteristics amoung bars. An example is given below.

Statistics Canada

All kinds of barplots suitable for categorical data may be used to represent quantitative data, but histograms cannot be used to represent categorical data.

Bias
Bias

The basic definition of statistical bias is as follows: the bias of an estimator (statistic) is how far the average statistic lies from the parameter it is estimating. In other words, if we imagine we could repeat a survey over and over again, and use the same method for each acquired sample to create the same statistic, then we expect the different values for the statistic to be randomly distributed around the parameter we are attempting to estimate. Bias occurs if those estimates for the statistic are systematically lower or systematically higher than the parameter value.

As an example: Let's say we wish to estimate the number of times a police officer requests a bribe in the course of a working day in a particular city. If we take a random sample of police officers from that city, and ask them directly how many bribes they request per working day, it is likely that the police officers will not be willing to tell us the true number of bribes requested. The estimate created from the data we collect will therefore be an underestimate; in other words, the estimate will be biased downward, or lower than the true average number of bribes requested by police officers in the course of a workday.

Let's say, instead, we pick a random sample of police officers and follow them around for a day, keeping track of the number of bribes we observe. It is likely that the police officers' behaviour will be different due to our observation; perhaps they ask for fewer bribes when watched. Again, then, the estimate created from the data we collect will be an underestimate.

Another way in which an estimate can end up biased is via data collected from a non-random subset of a population of interest. Returning to the example of the police officers, let's say we decide to follow a sample of police officers for a day, but pick those police officers by asking for volunteers. It is likely that the volunteer officers are fundamentally different from the non-volunteer officers; the volunteers may be less likely to bribe or behave inappropriately during the course of their duties. For that reason, the estimates we create from the data collected via the volunteer police officers will most likely be biased downward, as the worst of the bribers have little probability of entering our sample.

Finally, bias can occur because of bad questionnaire design, meaning bad question wording or some other systemic issue with a data-collection effort. For example, if we decide to interview the police officers of our city to determine their bribing propensity, but we word our question in a negative, condemning tone, they are less likely to be honest about their bribing than if we word the question in a neutral or supportive tone.

In summary, bias can be introduced into an estimate during any stage: developing a research project, picking a sample to survey, developing a questionnaire, or later in the process. Researchers must therefore be careful to consider sources of bias and do their best to mitigate those sources.

For more information about statistical bias, please see the following resources:

1. Wikipedia [disclaimer]

2. Statistics Glossary

3. Statistics Canada

4. A New View of Statistics

Bottom-up approach
Bottom-up Approach

bottom-up approach

democrative governance

Indigenous Peoples Rights Act (IPRA)

A bottom-up approach is not always necessary for conducting human rights and democractic governance measurement exercises; but it is valuable in providing reliable and useful results. Metagora advocates that in the field of democratic governance and human rights, sharing information is essential if it is to be efficiently used in policy-making. Indeed, one of the most frequently cited advantages of the bottom-up approach is that it promotes and increases “ownership” of these kinds of exercises.

The Metagora project shows that that bottom-up and top-down approaches can complement each other. For instance, a major benefit of global top-down approaches is that they facilitate comparisons, thus providing investors, donors and other bodies with indicators on governance. However, top-down approaches are relatively unsuited for monitoring and evaluating national and local policies or strategies aimed at improving human rights and democratic governance. Often excessively aggregated and generalised, top-down indicators fail to take into account the specific contexts of individual countries, and thus the information available does not provide relevant data upon which targeted policies can be created and developed. A major limitation of bottom-up approaches is precisely that they usually fail to facilitate international comparisons. Metagora was thus conceived as a necessary and useful complement to these top-down approaches, more than as a competing alternative.

Boxplot
Boxplot

Example: A Comparison of Statistical Estimates of Violations in Peru

Source: Science and Human Rights Program, American Association for the Advancement of Science.

A box plot is a graph that characterises the pattern of variation of the data. The plot simultaneously displays several measures of central tendency (what the “average” or “middle point” of the data is) and dispersion of the data (how spread out the data are).

The box plot provides the following information:

the position of the median;
the 25th and 75th percentiles (the 1st and 3d quartiles; these are the sides of the box); and
lines extending from the sides of the box as far as the minimum and maximum values.

Sometimes these lines will extend no further than one (or 1.5) inter-quartile range below the 1st quartile and above the 3rd quartile; in this case, points outside the lines will be individually identified.

CADT
See 'Certificate of Ancestral Domain Title' Below

Capacity Building
Capacity Building and
Skills Transfer

Capacity building

poverty

Skills transfer

There are many different types of capacity-building efforts, primarily determined by the community, group or organisation involved in the efforts. Those efforts include training, human-resource management, organisational development, the strengthening of communities and social networks, conducting research and coordinating alliances.

Capacity building is generally carried out by intergovernmental organisations, such as UNDP, private sector consulting firms and non-governmental organisations (NGOs). Irrespective of the processes and strategies used to build capacity, this term can be applied to interventions that have changed an organisation’s or community's ability to address development issues by creating new structures, approaches and/or values.

Cartesian Coordinate System
See 'X-Y Coordinate System' Below

Certificate of Ancestral Domain Title
Certificate of Ancestral Domain Title (CADT)

Certificate of Ancestral Domain Title (CADT)

indigenous peoples

indigenous peoples

CDHDF
Comisión de Derechos Humanos of Mexico City (CDHDF)

The CDHDF and its powers

The Mexico City Human Rights Commission (CDHDF) is the organisation in charge of investigating complaints and reports of alleged violations of human rights when they are attributed to any authority or public servant that has a job, position or commission in the pubic administration of Mexico City or in the law-enforcing organisations that have local jurisdiction in Mexico City.

The President of the CDHDF, also known as the Defender of the People, is assigned by the Mexico City Legislative Assembly (Asamblea Legislativa del Distrito Federal, ALDF), and his authority is autonomous, that is, it is not subject to any authority or public servant.

The CDHDF works according to its own laws and internal regulations. As stipulated in Article 17 of the Mexico City Human Rights Law, the powers of the CDHDF are to:

Receive complaints of alleged violations on human rights.
Know about and investigate written requests by the people of alleged violations on human rights in the following cases:
1. By acts or omissions of an administrative matter by public servants or by the Mexico City authorities, which third article of this Law refers to.
2. When a person or any social agent commits a crime with the consent of a public servant or local authority of Mexico City; or when the latter refuses, on no grounds, to exercise the obligations that they have in relation to these crimes, especially when they affect people’s physical integrity.
Make conciliatory proposals between the people who complain and the authorities or the public servants allegedly responsible, to immediately solve the problem when possible.
Make autonomous public recommendations, reports and complaints before the corresponding authorities.
Compel adherence to human rights in Mexico City.
Suggest to the different authorities in Mexico City, alterations in legislative decisions, and administrative practices that, according to the CDHDF, would lead to better protection of human rights.
Promote the study, teaching and promotion of human rights in its territorial area.
Issue its internal rules and regulations.
Make and orchestrate preventive programmes regarding human rights.
Ensure that the conditions of the people deprived of their freedom in detention, internment and social re-adaptation centres of Mexico City conform with the law, that the individuals’ human rights are fully respected, that medical examinations are performed on prisoners or those arrested when there are allegations of mistreatment or torture, and that the results of the revisions performed are reported to the appropriate authorities. These powers have no prejudice regarding the Human Rights National Commission (CNDH); coordination mechanisms will be used to exercise these powers.

Brief History of the CDHDF

Established on 30 September 1993, the Mexico City Human Rights Commission is the newest public organisation created to protect human rights in Mexico and is based on Article 102, section B of the Mexican Political Constitution.

The Defender of the People first appeared in Sweden at the beginning of the 19th century. The office exists now, in many variations, in many countries of the world. The Defender of the People is a mediator who seeks conciliatory ways to resolve conflicts. He/she is completely autonomous, and has the authority to resolve cases quickly, foregoing long judicial proceedings.

The People whom the CDHDF Serves

Any person that considers that his/her or another person’s human rights have been violated, regardless of his/her social condition, nationality, race, religion, sex, age, marital status, etc., can go to the Mexico City Human Rights Commission.

It is not a requirement to have a lawyer or an agent to make a complaint. A complainant only has to report, in either written or oral form, why he/she believes his/her rights were violated and present any evidence he/she might have. All information presented by the complainant is kept confidential.

All services are direct, free of charge, and operate on a 24h basis all year round.

Legal Authority of the CDHDF

(Article 18 and 19 of the Mexico City Human Rights Commission Law):

The Mexico City Human Rights Commission cannot become involved with cases regarding:

Acts and resolutions of electoral organisations or authorities;
Resolutions regarding jurisdictional matters;
Issues regarding working matters; and
Consultations posed by the authorities, citizens or other entities, on the interpretation of constitutional dispositions and of other juridical rules and regulations.

For the purposes of the above Law, jurisdictional resolutions are:

The final sentences that conclude the requests;
The interlocutory sentences that are brought in during the process;
The court orders and agreements passed by the judge, court staff, court or law-enforcing organisation, with a previous evaluation and juridical or legal process carried out before it is issued; and
Cases similar to those above concerning administrative issues.

All other acts or omissions of the procedures different to the ones mentioned in the previous sections can be considered to be administrative and, consequently, can be presented before the Mexico City Human Rights Commission.

The Commission does not examine jurisdictional matters.

For further information, please refer to www.cdhdf.org.mx/index.php?id=piwhr

CHRP
Commission on Human Rights of the Philippines (CHRP)

multistage cluster sampling

Protecting and promoting the human rights of all the people residing in the Philippines and Filipinos residing abroad, especially the underprivileged and disadvantaged sectors of society;
Engaging in sustained efforts with organisational integrity and competency in seeking justice; reorienting the agents of the State along human rights norms; advising the State on national policies and standards; catalysing effective and credible partnerships; and working with national and international organisations;
Advocating and monitoring the government's compliance with its international treaty obligations on human rights; and,
Encouraging civil society participation.

The organisation’s programmes and services are organised around four main axes:

Human Rights Protection Programme, which includes Investigation and Forensic Services, Legal Services, Legal Aid and Counseling, Conciliation, Mediation, Human Rights Assistance, and Visiting Services;
Human Rights Promotion Programme, which includes Education and Training, Human Rights Information and Public Advocacy, Human Rights Research and Development, Compliance Monitoring of International Human Rights Standards in Governance, and Web Services;
Human Rights Linkages Development and Strategic Planning, covering Rights-based Planning in Governance, Human Rights Performance System in Governance, Harnessing Competence in Government and None-state Actors for Good Human Rights Practices, the Executive Cooperation Programme, and the Legislative and Judicial Cooperation Programme; and,
Special Programmes, which include Rights-based Approach Application, Barangay Human Rights Action Centre, Human Rights Teaching Exemplars, Child Rights Centre, Women's Rights Programme Centre, and Asia-Pacific Institute of Human Rights.

Within the framework of the Metagora pilot project, the CHRP implemented a pilot survey focusing on the effective implementation of indigenous peoples’ human rights. In this exercise, the CHRP developed a strong collaboration with the Philippines National Statistical Coordination Board (NSCB) and the National Commission on Indigenous Peoples (NCIP). The activity consists of a small but incisive survey-based study implemented in three northern regions of the country with a high concentration of indigenous peoples. The objective of this pilot exercise was to develop evidence-based assessment methods and tools combining quantitative and qualitative approaches. The study aimed to measure four aspects of the rights of indigenous peoples to their ancestral domains and lands: the indigenous peoples’ perceptions and awareness of their rights, the enjoyment or violations of these rights, government measures and customary laws for the realisation of these rights, and the availability of mechanisms for redressing or fulfilling rights.

For further information regarding this pilot survey, please turn to the Synthesis Report produced by the Metagora Coordination Team.

Civil Rights
Civil Rights

Human rights

political rights

rights

the first generation refers to civil and political rights;
the second generation comprises economic, social and cultural rights; and,
the third generation refers to collective rights.

Civil rights are primarily designed to protect the individual against state interference, and are immediately applicable. They include the protection of life and security (i.e. the right to life, prohibition of torture and inhuman treatment or punishment, etc.); the prohibition of discrimination on any ground (such as race, sex, language, religion, political or other opinion, national or social origin, property, birth or other status); the protection of liberty (i.e., prohibition of arbitrary arrest and detention); and a set of freedoms (such as freedom of movement, of marriage, of religion, of peaceful assembly, of association, etc.). Civil rights can therefore be seen as the protections and privileges (rights and freedoms) that protect individuals from the state, thus ensuring their personal liberty.

Though distinct, civil rights and political rights are closely linked; the protection and fulfillment of one depends to a large extent on the fulfillment and protection of the other. Moreover, the distinction between civil and political rights is not always so obvious or clear; sometimes the two overlap.

All human rights are indivisible, interdependent and interrelated: the fulfillment and protection of civil and political rights depend on, and are required for, other categories of human rights.

In international human rights law, civil rights are essentially protected by the International Covenant on Civil and Political Rights (ICCPR), which was drafted in 1966 and entered into force in 1976. Adherence to the Covenant is monitored by the Human Rights Committee. Over time, additional protocols and instruments were created which also aim to protect civil rights.

All States Parties to the Covenant are required to submit regular reports to the Committee on how they are implementing civil and political rights. Such information is provided by self-reporting and, thus, can be limited. The reports provided are examined by the Committee which is composed of independent experts appointed by the United Nations. The Committee then addresses its concerns and recommendations in the form of "concluding observations." In the First Optional Protocol to the Covenant, the Committee was given jurisdiction to examine individual complaints; this is not yet the case with the Committee established to monitor economic, social and cultural rights [1].

1. For further information, please refer to the web site of the OHCHR, http://www.ohchr.org/english/bodies/hrc/index.htm.

Closed-ended Question
Closed-ended versus Open-ended Questions [1]

Closed-ended questions limit respondents' answers to the survey. The participants are allowed to choose from either a pre-existing set of dichotomous answers, such as yes/no, true/false, or multiple choice with an option for "other" to be filled in, or ranking-scale response options. For example, a closed-ended question might ask for a respondent's religion, giving several religion categories (i.e., Catholic, Protestant, Buddhist, Muslim, etc.) and an "other" option.

The most common of the ranking-scale questions is called the Likert scale question. This kind of question asks the respondents to look at a statement (such as "The most important education issue facing our nation in the year 2000 is that all students in their third year of primary school should be able to read") and then "rank" this statement according to the degree to which they agree ("I strongly agree, I somewhat agree, I have no opinion, I somewhat disagree, I strongly disagree").

The advantages of closed-ended questions are:

Closed-ended questions are more easily analysed. Every answer can be given a number or value so that a statistical interpretation can be assessed. Closed-ended questions are also better suited for computer analysis. If open-ended questions are analysed quantitatively, the qualitative information is reduced to coding and answers tend to lose some of their initial meaning. Because of the simplicity of closed-ended questions, this kind of loss is not a problem.
Closed-ended questions can be more specific, thus more likely to communicate similar meanings. Because open-ended questions allow respondents to use their own words, it is difficult to compare the meanings of the responses.
In large-scale surveys, closed-ended questions take less time from the interviewer, the participant and the researcher, and so they are a less expensive survey method. Generally, the response rate is higher with surveys that use closed-ended question than with those that use open-ended questions.

A limitation of closed-ended questions is the assumption that the researcher knows enough about the phenomenon being studied and about the respondents' perceptions to be able to build an appropriate and sensitive set of categories. If that is not true, the responses might be grouped into inappropriate categories or concepts. When using closed-ended questions, the researcher might first have an exploratory survey during which a small sample is asked open-ended questions. The answers obtained can be used to form categories and/or check the researcher's assumptions.

Open-ended questions do not give respondents answers to choose from, but rather are phrased so that the respondents are encouraged to explain their answers and reactions to the question with a sentence, a paragraph, or even a page or more, depending on the survey. If you wish to find information on the same topic as asked above (the future of elementary education), but would like to find out what respondents would come up with on their own, you might choose an open-ended question like "What do you think is the most important educational issue facing our nation in the year 2000?" rather than use the Likert scale question. Or, if you would like to focus on reading as the topic, but would still not like to limit the participants' responses, you might pose the question this way: "Do you think that the most important issue facing education is literacy? Explain your answer below."

The advantages of open-ended questions are:

Open-ended questions allow respondents to include more information, including feelings, attitudes and understanding of the subject. This allows researchers to better access the respondents' true feelings on an issue. Closed-ended questions, because of the simplicity and limit of the answers, may not offer the respondents choices that actually reflect their real feelings. Closed-ended questions also do not allow the respondents to explain that they do not understand the question or do not have an opinion on the issue.
Open-ended questions cut down on two types of response error: respondents are not likely to forget the answers they have to choose from if they are given the chance to respond freely; and open-ended questions simply do not allow respondents to disregard reading the questions and just "fill in" the survey with all the same answers (such as filling in the "no" box on every question).
Because they can elicit extra information from the respondent, such as demographic information (current employment, age, gender, etc.), surveys that use open-ended questions can be used more readily for secondary analysis by other researchers than can surveys that do not provide contextual information about the survey population.
Research has shown that open-ended questions are better for eliciting sensitive information, such as information about sexual assault or drug usage, than closed-ended questions.

Note: Keep in mind that you do not have to use closed-ended or open-ended questions exclusively. Many researchers use a combination of closed and open questions; often researchers use closed-ended questions in the beginning of their survey, then allow for more expansive answers once the respondent has some background on the issue and is "warmed-up."

1. This definition is modified from Types of Questions (accessed 28 December 2006).

Cluster Sampling
Cluster Sampling

one-stage cluster sampling

two-stage cluster sampling

In cluster sampling, the clusters are the primary sampling unit (PSU’s) and the units within the clusters are the secondary sampling units (SSU’s). It is important to keep these two levels in mind when calculating standard errors from cluster samples. If a cluster sample is analysed as if it were a simple random sample, the reported standard errors would probably be smaller then they should be. That would give the impression that the survey results are more precise than they really are. Whereas stratification often increases precision of the estimation compared with simple random sampling, cluster sampling often decreases it. That is because units in a cluster tend to be more similar than elements selected at random from the whole population. When using cluster sampling, it is usually necessary to increase the total sample size to achieve the same precision as in simple random sampling. Nevertheless, there are cases where cluster sampling is useful.

The main reason for using cluster sampling is that it usually much cheaper and more convenient to sample the population in clusters rather than randomly. In some cases, constructing a sampling frame that identifies every population element is too expensive or impossible. Cluster sampling can also reduce cost when the population elements are scattered over a wide area. Suppose you want to survey school children of a certain age in a specific area. If you drew a simple random sampling of school children, you might have to visit all schools in the area to interview your sample. With cluster sampling you could first select the schools to be included in your sample, and then select school children within each of the selected schools. That would probably reduce the number of schools you have to visit and therefore reduce the cost of data collection. In this example, the schools are what are sometimes referred to as natural clusters. In other cases, the population may be widely distributed geographically, and then cluster sampling, where the clusters consists of geographical areas, could reduce the number of areas that need to be visited. A smaller number of areas that need to be visited could reduce travel expenses and also make possible more efficient supervision of the fieldwork.

For more information about cluster sampling, see: Sarndal, C.E., Swenson, B., and Wreman, J.H., Model Assisted Survey Sampling, Springer-Verlag, New York, 1992.

Coding
Coding

controlled vocabulary

Human Rights Data Analysis Group (HRDAG)

Coding is a necessary step in the development of statistics from qualitative data sources. Coding is also sometimes used in the field during a survey, for collecting quantitative information. In that case, the controlled vocabulary/data categories are determined prior to data collection, and the interviewer is responsible for coding a respondent's qualitative information and recording the appropriate code on the questionnaire.

Cognitive Interviewing
Cognitive Interviewing [1]

The cognitive interviewing approach to evaluating sources of response error in survey questionnaires was developed during the 1980s through an interdisciplinary effort by survey methodologists and psychologists. It explicitly focuses on the cognitive processes that respondents use to answer survey questions; therefore, processes that are normally hidden, as well as observable processes, are studied. Although cognitive interviewing is a powerful tool, cognitive interviewing is practiced in relatively few places, mostly in federal statistical agencies and survey research organisations in North America and Europe.

There are two major sub-types of cognitive interviewing methods: think-aloud interviewing and verbal probing techniques. The term think-aloud is used here to describe a specific type of activity in which subjects are explicitly instructed to "think aloud" as they answer the survey questions. The interviewer reads each question to the subject, and then records and/or otherwise notes the processes that subject uses in arriving at an answer to the question. The interviewer interjects little else, except to say "tell me what you're thinking" when the subject pauses. As an alternative to the think-aloud, verbal probing is used. After the interviewer asks the survey question, and the subject answers, the interviewer then asks for other, specific information relevant to the question, or to the specific answer given. In general, the interviewer "probes" further into the basis for the response.

The two general approaches to probing are: concurrent probing and retrospective probing. With concurrent probing, the interchange is as follows: a) the interviewer asks the survey question; b) the subject answers the question; c) the interviewer asks a probe question; d) the subject answers the probe question; and e) possibly, further cycles of (c-d). In retrospective probing, the subject is asked the probe questions after the entire interview has been administered, sometimes in a separate part of the interview known as a “debriefing session.”

Cognitive interviewing can take place either in a particular "laboratory" or in location(s) similar to those used during the actual survey interviews. Although organisations that conduct relatively large numbers of cognitive interviews have dedicated laboratory facilities containing video and audio equipment, and remote observation capability, cognitive interviewing does not require special physical environments, or sophisticated recording equipment. Equipment needs are minimal. It is helpful to have a tape-recorder, as it is useful to record interviews (most subjects do not object, as long as privacy and confidentiality requirements are met).

1. This definition is adapted from Willis, G., Cognitive Interviewing: A How-To Guide, 1999. Further information on cognitive interviewing can be found in that resource.

Comisión de Derechos Humanos of Mexico City
See 'CDHDF' Above

Commission on Human Rights of the Philippines
See 'CHR' Above

Complex Indicator
See 'Indicator' Below

Complex Sample Design
Sample Design

A complex sample design may differ from the simple procedure in one or several of three different ways: unequal probabilities, stratification, or multi-stage sampling.

Unequal probabilities may be used to over- or under-represent a specific sub-population in the sample (for example, to have a proportion of youth or of unemployed different from the corresponding proportion in the population). An indirect way of sampling can also over-represent certain individuals: for example, a simple random sample of pupils is selected in the schools of a city where their brothers and sisters will also be interviewed. In that case, a four-child family has a probability for being selected that is four times greater than that of a single-child-family. In all cases, correct estimates imply that the answers be weighted by the inverse of the probabilities the different observation units had to be in the sample.

Some multi-stage sampling uses probability-proportional-to-size selection. In this method, in order to have an equal final probability for all individuals, the same number of individuals is selected in every primary unit, so that the probability of selecting an element varies inversely with the size of the primary unit. The final probability is then the product of the probabilities at the two stages, and hence is the same throughout the sample.

Stratification selects independent samples for different sub-populations. This often allows a reduction of the sampling error. Multi-stage sampling considers two or more levels of units imbedded one in the other, such as geographic areas (primary units), factories (secondary units), employees (tertiary units). At each stage, a sample of the corresponding units is selected. A multi-stage procedure may, at any of its different stages, be combined with stratification or/and with unequal probabilities.

Confidence Interval
Confidence Interval

(mean – 1.96*se, mean + 1.96*se)

Changing the constant (i.e., the "1.96") changes the percentage associated with the confidence interval; for example, a 90 percent confidence interval for a mean created from a simple random sample is:

(mean – 1.645*se, mean + 1.645*se)

Confidence intervals can be created for other statistics as well, but creating those confidence intervals can be more difficult [1].

The form given above for a confidence interval is not complicated in itself; what makes confidence intervals complicated to interpret are two factors:

The assumptions that must be valid in order for the confidence interval to be valid; and
Understanding the proper interpretation of a confidence interval.

There are two basic assumptions that must be true for a confidence interval to "work." First, the data used to develop the statistic for which the confidence interval is created must be mostly "symmetric," meaning, they can't have too much skew. Second, there must be enough data used to develop the statistic for which the confidence interval is created. Usually, there should be at least 30 data points available. These two assumptions interact when we are creating a confidence interval for a mean, in that the more data that are collected/available, the more skew we can tolerate in those data.

If these assumptions are not valid, meaning, if the data have too much skew or there aren't enough data, then confidence intervals derived using the equations above may not be meaningful.

Interpreting the meaning of a confidence interval can be tricky. What is random about a confidence interval is the endpoints of that interval.

For example, let's assume that we are to take a simple random sample of 100 members of a population for the purpose of calculating the mean height (in centimeters) of members of that population. Imagine that we could repeat our experiment 100 times, that is, collect 100 samples and, for each sample, calculate a mean and a standard error. We would be able to form a confidence interval for each of our 100 samples.

The 95 percent confidence interval is then defined as the interval such that in approximately 95 of the experiments, the parameter (true mean height) we are estimating is “captured” within the confidence interval, and in approximately five of the experiments, the parameter (true mean height) we are estimating is not “captured” within the confidence interval. Of course, we don’t repeat our experiment 100 times; we just do it once. What we are counting on is that the experiment we do is not one of the approximately five for which the confidence interval does not capture the parameter.

The illustration below demonstrates the results of repeating this experiment 50 times. The mu in the graphic represents the parameter we are estimating, that is, the true mean height of member of the population of interest. Each line represents one experiment, that is, the confidence interval calculated from one simple random sample of 100 members of the population. Note that in three cases, the 95 percent confidence interval does not contain mu, which is about what we would expect after 50 repeats of the experiment (about 5 percent of the cases).

Taken from http://en.wikipedia.org/wiki/Confidence_interval

Source: Wikipedia [disclaimer]

1. For example, the confidence interval for a median is calculated as given by www.umanitoba.ca/centres/mchp/concept/dict/ (26 December 2006), and the confidence interval for a ratio of means is calculated as given by www.graphpad.com/FAQ/images/Ci of quotient.pdf (26 December 2006).

Controlled Vocabulary
Controlled Vocabulary [1]

coding

In order to create a controlled vocabulary and ensure the quality of the data, every definition within the controlled vocabulary must satisfy the following five properties:

Mutually exclusive: No single item to be coded can fit into more than one definition in the controlled vocabulary.
Exhaustive: A definition must exist for every possible item that can occur in the qualitative data being studied.
Distinguished: Each definition must have an explicit characteristic that distinguishes it from all others in the controlled vocabulary.
Exemplified: Each definition must be accompanied by examples showing how to apply the definition in a specific situation.
Countable: Each definition must contain a counting rule explicitly stating how items are enumerated.

1. This definition is based on the Controlled Vocabulary definition given on the HRDAG Web site.

Correlation
Correlation [1]

Correlation

Although the correlation described above is fairly obvious, your data may contain unsuspected correlations. You may also suspect there are correlations, but don't know which are the strongest. An intelligent correlation analysis can lead to a greater understanding of your data.

Like all statistical techniques, correlation is only appropriate for certain kinds of data. Correlation works for data in which numbers are meaningful, usually quantities of some sort. It cannot be used for purely categorical data, such as gender, brands purchased or favourite colour.

The main measure of correlation in a dataset is a statistic called the correlation coefficient (or r). It ranges from -1 to +1. The closer r is to +1 or -1, the more closely the two population characteristics are related. If r is close to 0, it means there is no relationship between the population characteristics. If r is positive, it means that as one population characteristic gets larger the other gets larger. If r is negative it means that as one gets larger, the other gets smaller (often called an "inverse" correlation).

Never assume a correlation means that a change in one population characteristic causes a change in another. Sales of personal computers and athletic shoes have both risen strongly in the last several years and there is a high correlation between them, but you cannot assume that buying computers causes people to buy athletic shoes, or vice versa.

The second caveat is that the correlation statistic is only meaningful with respect to linear relationships: as one population characteristic gets larger, the other gets larger (or smaller) in direct proportion. It does not work well with curvilinear relationships, in which the relationship does not follow a straight line. An example of a curvilinear relationship is age and health care. They are related, but the relationship doesn't follow a straight line. Young children and older people both tend to use much more health care than teenagers or young adults. Multiple linear regression can be used to examine curvilinear relationships, but it is beyond the scope of this encyclopedia.

1. This definition is based on www.surveysystem.com/correlation.htm (accessed December 28 2006).

Cultural Rights
Cultural Rights

the first generation refers to civil and political rights;
the second generation comprises economic, social and cultural rights; and,
the third generation refers to collective rights.

There is no simple definition of the right to culture, though the right to express and enjoy one’s culture does exist. International legislation recognises, as cultural rights, the right to take part in cultural life; the right to enjoy the benefits of scientific progress and its applications; the right to benefit from the protection of the moral and material interests resulting from any scientific, literary or artistic production of which he/she is the author; etc.

Contrary to civil and political rights, which are immediately applicable and essentially based on the prohibition of States to do something (i.e., resort to torture, take actions that curtail freedom of speech, freedom of religion, or the right to vote, etc.), cultural rights tend to be considered as requiring States to take active and specific measures, such as legislation, policies or programmes, so that those rights can be realised. Their realisation is seen as progressive: “full economic, social, and cultural rights can be achieved only gradually. Resources and time may be required” [1], though it is also clearly stated that full rights should be reached over time, and that States have a legal obligation to take immediate and continued action to do so. Moreover, any action, whether legal or political, taken to diminish existing protections and levels of realisation of these rights should be prohibited.

All human rights are indivisible, interdependent and interrelated: the fulfillment of one right affects that of others. This is true in general as well as specifically, that is among all rights, and among or within categories of rights. Cultural rights are closely linked to social and economic rights, and the difference among all three is not always obvious. For instance, the right to education has been considered by different experts as a social, economic or cultural right.

In international human rights law, cultural rights are essentially protected by the International Covenant on Economic, Social and Cultural Rights (CESCR). This Covenant is monitored by the Committee on Economic, Social and Cultural Rights, which is composed of independent experts appointed by the United Nations. The Committee is responsible for monitoring the implementation of the Covenant by its States Parties. These States are required to submit regular reports on how they are implementing these rights. It should be stressed here that such information is provided through self-reporting and thus can be limited. The reports provided are examined by the Committee, which then elaborates “concluding observations” in which it addresses its concerns and recommendations. To date, the Committee is not enabled to consider individual complaints against States Parties, though a draft Optional Protocol, under consideration, could provide the Committee with the jurisdiction to do so [2].

1. McChesney, A., Promoting and Defending Economic, Social and Cultural rights, AAAS/HURIDOCS, Washington, DC, 2000, p. 18.

2. For further information, please refer to the web site of the OHCHR, www.ohchr.org/english/bodies/cescr/index.htm.

Data
The Spectrum from Qualitative to Quantitative Data

Data

There are three categories of data used to create statistics and indicators to inform policy processes: story data, categorical data, and numeric data.

Story data are data in the form of stories, statements, written documents, or other non-numeric forms. Such data are the most "qualitative" of the three categories of data given above, and qualitative methods must be used to analyze them. However, it is possible to convert story data into categorical data through a controlled vocabulary and coding procedure.

Categorical data are data that take a finite set of values that can be either numeric or categorical. For example, eye color, ethnicity, and country of residence are all possible types of categorical data that are not numeric (the categories can not be ordered meaningfully from a "lesser" to a "greater" value). Statisticians often refer to categorical data as "qualitative data," while social scientists might consider categorical data to be "quantitative."

Numeric data are data that exist in numeric form, such as height, the number of children in a household, and annual income. Numeric data is most definitely quantitative, and can be either “continuous,” such as time, or income, or “discrete,” such as number of children. It is almost always possible to convert numeric data into categorical data. For example, the number of children in a household can be binned into three categories: none, 1-4, or 5 or more.

Both categorical data and numeric data can be used to create statistics.

In analysing a complex concept or situation, a combination of story data and categorical/numeric data can be more effective than either by itself. While categorical and numeric data tend to illustrate and summarise overall trends and patterns, story data can add detail to the "picture" and explain the causes of some of those trends and patterns. In that way, a researcher can view both the "forest," the overall trends given by the categorical and numeric data, and the "trees," the details and richness provided by the story data, and create the best possible analysis.

Decile
See 'Percentile' Below

Democracy
Democracy

demos

kratos

governance

The notion of democracy, far from being uniform and universally accepted, covers a variety of models. In broad terms, democracy can be seen as falling into two basic categories: direct and representative. In a direct democracy, all citizens participate directly in public decisions; in a representative democracy, citizens elect officials to make political decisions, formulate laws, and administer programmes for the public good. Obviously, models of direct democracy are only practical with relatively small numbers of people and, in modern societies, most democracies are representative, though the way in which officials are elected can vary enormously.

As noted by the International Institute for Democracy and Electoral Assistance (IDEA), different definitions and theories of democracy emphasise different aspects. IDEA’s understanding of democracy is “that of a system of political governance whose decision-making power is subject to the controlling influence of citizens who are considered political equals. A democratic political system is inclusive, participatory, representative, accountable, transparent and responsive to citizens’ aspirations and expectations” [1]. Moreover, democracy is a process, "not an all-or-nothing affair. It is a question of the degree to which citizens exercise control over political decision-making and are treated as equals”.

Over the years, IDEA has developed a methodology for assessing democracy. The assessment framework is based on the belief that democracy should be defined by its basic principles or values. These include “popular control over public decision-making and decision-makers, and equality between citizens in the exercise of that control” [2]. The belief behind assessment exercises is that governing arrangements and procedures can be considered democratic if these principles are respected and implemented. These principles are further defined as covering notions of: “participation, authorisation, representativeness, accountability, transparency, responsiveness and solidarity” [3].

Although the word “democracy” does not appear in the United Nations Charter, and the right to democracy is not explicitly codified in human rights law, the latter includes many references to fundamental principles of democracy. For instance, the Universal Declaration of Human Rights indicates and refers to one of the central principles of democracy in its Article 21, para. 3: “The will of the peoples shall be the basis of the authority of government; this will shall be expressed in periodic and genuine elections which shall be by universal and equal suffrage and shall be held by secret vote or by equivalent free voting procedures.” In addition, the Universal Declaration of Human Rights refers to a series of rights that are directly linked to the concept of democracy, including the right to take part in government and the right to equal access to public services [4]. The 1993 World Conference on Human Rights, held in Vienna, concluded that democracy, development and respect for human rights are interdependent and mutually reinforcing [5]. In recent years, several United Nations documents, such as the United Nations Millennium Declaration, adopted in 2000, have further stressed the interdependence of democracy and human rights.

1. For further information, please consult IDEA’s website at http://www.idea.int/democracy/index.cfm

2. See IDEA, Democracy Assessment: The basics of the international IDEA assessment framework.

3. Ibid.

4. See See Symonides, J., and Volodin, V. (eds), A Guide to Human Rights: Institutions, Standards and Procedures, UNESCO, Paris, 2001, p. 368.

5. See http://www.unhchr.ch/democracy/ and The Vienna Declaration and Programme of Action, 1993.

Descriptive Statistics
Descriptive Statistics and
Inferential Statistics [1]

Descriptive statistics

central tendency

dispersion

Inferential statistics, or statistical induction, means the use of statistics to make inferences concerning some unknown aspect of a population from a sample of that population. A common method used in inferential statistics is estimation. In estimation, the sample is used to estimate a parameter, and a confidence interval about the estimate is constructed. Other examples of inferential statistics methods include hypothesis testing, linear regression, and principle components analysis.

1. This definition is based in part on http://en.wikipedia.org/wiki/Inferential_statistics [disclaimer] and http://davidmlane.com/hyperstat/A29136.html.

Development Aggression
Development Aggression

bottom-up

Développement, Institutions & Analyses de Long Terme
See 'DIAL' Below

DIAL
Développement, Institutions & Analyses de Long Terme (DIAL)

policies

DIAL is the result of a partnership between the Institute for Development Research (IRD) and the French Development Agency (AFD). It is funded by the Ministry of Foreign Affairs (General Direction of International Cooperation and Development) and the Ministry of Economy and Finances (National Institute of Statistics and Economic Studies). DIAL includes a research unit of IRD, with research branches in Senegal (Dakar University) and soon also in Bamako, Mali, and Hanoi, Vietnam.

DIAL is composed of a multi-disciplinary team of economists, demographers, and statisticians who are investigating the links between the demo-economic development process, resource distribution and state intervention. In the research process, production of statistics on priority issues, in-depth analysis of the collected data, and evaluation of economic policies are closely associated.

The institution’s current research programme (2005-2008) follows three principal areas:

Genesis of inequalities and poverty, including the dynamics of poverty, job market and social mobility, rural dynamics, multiple dimensions of poverty, and the impact of economic policies on distribution;
o Governance, institutions and long-term distribution of resources, including participation, identities and construction of the State, and a historical perspective of resource distribution in Africa. In this context, within the framework of Metagora, DIAL assists with and coordinates the implementation of modules on governance, democracy, and subjective poverty in francophone Africa and in the Andean Community; and,
Aid, international strategies and global inequalities, including development aid policies, and monitoring and evaluating new international strategies in the fight against poverty.

The DIAL has been involved in the Metagora pilot project through the implementation of two multi-country regional activities, conducted on the basis of the 1-2-3 survey methodology that they have developed over the years. Thirteen National Statistical Agencies were involved in French-speaking Africa and in the Andean region. These regional activities explore the possibilities of using official household surveys as a tool for measuring issues related to governance, subjective poverty, and democratic participation.

For further information regarding DIAL or its participation in the Metagora project, please consult the web site of the organisation at www.dial.prd.fr or the Metagora Synthesis Report.

Discrimination
Discrimination

human right

ethnic

The 1945 Charter of the United Nations, along with the Universal Declaration of Human Rights (1948) and the two 1966 United Nations Human Rights Covenants guarantee respect for human rights without distinction of any kind.

Article 2 of the Universal Declaration of Human Rights states that “Everyone is entitled to all the rights and freedoms set forth in this Declaration, without distinction of any kind, such as race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth or other status.”
Article 26 of the Covenant on Civil and Political Rights also stresses that “All persons are equal before the law and are entitled without any discrimination to the equal protection of the law. In this respect, the law shall prohibit any discrimination and guarantee to all persons equal and effective protection against discrimination on any ground such as race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth or other status.”

The prohibition of discrimination is therefore violated in cases where a disadvantageous distinction is made using one of the attributes listed above. Attributes such as age, health, disability or sexual orientation may also amount to discrimination.

Over time and within the framework of the UN, several treaties have been created to address specific categories of discrimination such as the 1965 Convention on the Elimination of All Forms of Racial Discrimination and the 1979 Convention on the Elimination of All Forms of Discrimination against Women. In 1989, the Human Rights Committee defined the term "discrimination" as follows: “any distinction, exclusion, restriction, or preference which is based on any ground such as race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth, or other status, and which has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise by all persons on an equal footing, of all rights and freedoms” [1].

Equal treatment of persons is thus an essential and central element of international human rights law. The prohibition of discrimination implies different elements, depending on the actual situation. Indeed, as mentioned by Walter Kälin, Lars Müller and Juditu Wyttenbach, [2] discrimination, and therefore a situation of human rights violations, can be embodied by: “unequal treatment by the law of persons who find themselves in the same situation”; “unequal application of legal provisions on such persons without well-founded, objective reasons”; or “equal treatment of persons who find themselves in different situations justifying differential treatment” [3].

1. See Symonides, J. and Volodin, V. (eds), A Guide to Human Rights: Institutions, Standards and Procedures, UNESCO, Paris, 2001, p. 161.

2. See Kälin, W., Müller, L., and Wyttenbach, J., The Face of Human Rights, Lars Müller Publishers, Switzerland, 2004, p. 120.

3. Ibid.

Distrito Federal
Distrito Federal (Mexico City)

The core of the vast Mexico City Metropolitan Area is geographically situated within the limits of the Federal District (D.F.), though, as such, it is bigger than the D.F. The Metropolitan Area includes the entire D.F. together with some municipalities of the Estado de Mexico and of the state of Hidalgo.

The D.F is composed of 16 boroughs: Álvaro Obregón, Azcapotzalco, Benito Juárez, Coyoacán, Cuajimalpa de Morelos, Cuauhtémoc, Gustavo A. Madero, Iztacalco, Iztapalapa, Magdalena Contreras, Miguel Hidalgo, Milpa Alta, Tláhuac, Tlalpan, Venustiano Carranza, and Xochimilco.

Economic Rights
Economic Rights

the first generation refers to civil and political rights;
the second generation comprises economic, social and cultural rights; and,
the third generation refers to collective rights.

Economic rights include the right to work, the right to the free choice of employment and to just and favourable conditions of work; the right to form and join trade unions: the right to strike; the right to social security; and the right to own property.

Contrary to civil and political rights, which are immediately applicable and essentially based on the prohibition of States from doing something (i.e., resort to torture, take actions that curtail freedom of speech, freedom of religion, or the right to vote, etc.), economic rights tend to be considered as requiring States to take action, usually in the form of specific legislation, policies or programmes, so those rights can be realised. The realisation of those rights is seen as progressive: “full economic, social, and cultural rights can be achieved only gradually. Resources and time may be required” [1], though international legislation clearly states that full rights should be reached over time, and that States have a legal obligation to take immediate and continued action to do so. Moreover, any action, whether legal or political, taken to diminish existing protections and levels of realisation of these rights should be prohibited.

All human rights are indivisible, interdependent and interrelated, and the fulfillment and protection of one right affects that of others. This is true among all rights and among or within specific categories of rights. For instance, economic rights are closely linked to social and cultural rights. The right to work, for example, is connected to that of ensuring minimum standards of living, etc. Just as the distinction between civil and political rights is sometimes blurred, the difference between economic, social, and cultural rights is not always obvious. For example, the right to education has been considered by different experts as an economic, social or cultural right.

In international human rights law, the realisation of economic rights is provided for in Chapter IX of the UN Charter and in the Universal Declaration of Human Rights, the International Covenant on Economic, Social and Cultural Rights (CESCR), the International Labour Organisation, and various regional documents.

The CESCR is monitored by the Committee on Economic, Social and Cultural Rights, which is composed of independent experts appointed by the United Nations. The Committee is responsible for monitoring the implementation of the Covenant by its States Parties. They are required to submit regular reports on how they are implementing these rights. Such information is provided through self-reporting and thus may be limited. The reports provided are examined by the Committee, which then elaborates “concluding observations” in which it addresses its potential concerns and recommendations. To date, the Committee is not enabled to consider individual complaints against States Parties, though a draft Optional Protocol, under consideration, could provide the Committee with the jurisdiction to do so [2].

1. McChesney, A., Promoting and Defending Economic, Social and Cultural rights, AAAS/HURIDOCS, Washington DC, 2000, p. 18.

2. For further information, please refer to the website of the OHCHR, http://www.ohchr.org/english/bodies/cescr/index.htm

Enumerator
See 'Interviewer' Below

Ethnicity
Ethnicity [1]

ethnic group

ethnogenesis

In terms of demographic data collection, ethnicity is defined differently by different government entities. Two examples of government definitions of ethnicity are presented below.

In the United States statistical system, ethnicity is only defined in terms of "Hispanic," meaning the various Latin American ethnic groups plus the Spanish, and "non-Hispanic" (everyone else). Race is defined separately. Data on ancestry, a person’s origin or descent, “roots,” or heritage, or the place of birth of the person or the person’s parents or ancestors before their arrival in the United States, are collected. In the United States, it is illegal to discriminate against a person based on their ethnicity, race, or ancestry.

In contrast, the People's Republic of China has officially split the population into 56 ethnic groups, of which the most numerous are the Han Chinese. Many of the ethnic minorities maintain their own individual culture and language, although many are also becoming more like the Han. The Han Chinese are the only ethnic group bound by the one-child policy and many villages falsified a change in their ethnic group, for example from Han to Manchu, to avoid the policy.

There is a degree of autonomy granted to areas with a high minority population. Inner Mongolia is an example of such. Sometimes ethnic minorities are allowed to use their own language in official documents, but not always. For example, a Tibetan can request an official document to be in either the Chinese or Tibetan language. But a Han Chinese can only request Chinese. Some ethnic groups do not have this option, such as the Hui, who can only request Chinese.

There is no equal opportunity law in China, and although the ethnic groups are said to be equal, it is commonplace to specify which ethnic group is preferred, or even required, when, for example, advertising employment. Most official government bodies are required to employ at least one member of an ethnic minority. Sometimes people are given the choice of which ethnic group they wish to belong to, but "mixed-race" is not an option.

As can be seen from the examples above, collecting data on ethnicity is a double-edged sword. On the one hand, information on ethnicity can be used to guard against discrimination against ethnic minorities and increase the visibility of ethnic minorities within the government. On the other hand, data on ethnicity can be used to perpetuate human rights abuses against ethnic minorities. The decision to collect ethnicity data in order to inform a particular public policy must be made only after serious consideration of the ethical implications of the data-collection project.

1. This definition is essentially based on http://en.wikipedia.org/wiki/Ethnic_group (accessed 27 December 2006) [disclaimer].

Ethnographic Study
Ethnographic Study

Expert Interviews (Pre-Questionnaire Design)
Expert Interviews (Local Consultations, Exploratory in-depth Interviews)

Expert interviews, or local consultations, may take place before a questionnaire is designed so that:

appropriate vocabulary is used to describe the experiences and concepts about which the respondents will be queried;
sensitivities towards the topics covered by the questionnaire can be gauged;
the universe of experiences related to the topic of interest can be better defined; and
ethical issues that may arise during the data-collection process can be explored.

"Experts" need not be academic experts; they might be cultural experts who are familiar with the concepts to be studied but might have very little formal education on those concepts.

Expert interviews are not designed to gather quantitative information, and are not usually limited by a particular set of questions. Rather, they are considered qualitative fact-finding missions, during which new questions may arise. Such interviews take at least a few hours, and usually a limited number of expert interviews are planned. When the expert interviews are completed, the qualitative data are summarised and studied, and the design of the questionnaire can begin.

Expert Review
Expert Review

Expert review

questionnaire design

cognitive interviewing

questionnaire

Subject-matter Experts: Even if the questionnaire designer is an expert in the subject on which the questionnaire has been written, he/she should consult with other experts on whether the questions, as designed, will capture the desired information accurately. For example, a subject-matter expert may find that a closed-ended question does not contain enough or appropriate answer categories, or that an important aspect of the subject is not covered adequately by the questionnaire.
Cultural Experts: These types of experts are especially important for a survey that is to be administered to respondents from multiple cultures, or from a culture with which the questionnaire designer is not intimately familiar. Cultural experts can pick up on wording that may have a different meaning in one culture than the next, point out topics that are taboo or sensitive, and discuss access to certain subpopulations in the culture, such as women.
Questionnaire-design Experts: Even experts in questionnaire design consult with other experts in questionnaire design during the development of a survey instrument. Such experts are trained to look for vague wording, double-barreled questions, problems with skip patterns through the survey, misleading questions, and questions with multiple interpretations.

Once a questionnaire has been modified on the basis of the expert comments, it will be ready for the formal testing phase of questionnaire design.

Exploratory Data Analysis
Exploratory Data Analysis [1]

descriptive statistics

The particular graphical techniques employed in EDA are often quite simple, consisting of various techniques of:

Plotting the raw data, using tools such as line plots, histograms, scatterplots, age-sex pyramids, barplots, and stem and leaf plots;
Plotting simple statistics using methods such as box plots;
Positioning such plots so as to maximize our natural pattern-recognition abilities, such as using multiple plots per page; and
Exploring other statistics, such as means, medians, modes, standard deviations, ranges, and percentiles.

In his seminal work on the subject, Exploratory Data Analysis (Addison-Wesley, 1977), John Tukey suggests that one thinks of exploratory analysis as the first step in a two-step process similar to that used in criminal investigations. In the first step, one searches for evidence using all of the investigative tools that are available. In the second step, that of confirmatory data analysis, one evaluates the strength of the evidence and judges its merits and applicability. It is in this second step that one would likely apply the techniques of inferential statistics.

1. This definition is based on www.itl.nist.gov/div898/handbook/toolaids/pff/1-eda.pdf, www.datamology.com/eda.shtml, and www.statistics4u.info/fundstat_eng/.

Exploratory in-depth Interviews (Pre-Questionnaire Design)
See 'Expert Review' Above

Family
Families, Households, and Housing Units

Source: http://ca.rand.org/stats/census/censusdefs.html (26 December 2006)

Family A householder and one or more people living in the same household who are related to the householder by birth, marriage, or adoption.

Household Includes all of the people who occupy a housing unit.

Housing Unit A house, an apartment, a mobile home, a group of rooms, or a single room that is occupied as separate living quarters.

There are clear issues and ambiguities in those definitions, however. What about people living in group quarters, such as a dormitory: are those people part of one household? What about a child who travels during the week to attend school and then returns to her "family" for the weekend: in what household does that child reside? What if a single housing unit is occupied by more than one family because of a humanitarian emergency? How do the definitions given above translate to a refugee camp? To a polygamous society in which each wife lives in a separate housing unit?

The United Nations gives a different, and more open, set of guidelines for defining families, households, and housing units, as shown below:

Family "The family within the household, a concept of particular interest, is defined as those members of the household who are related, to a specified degree, through blood, adoption or marriage. The degree of relationship used in determining the limits of the family in this sense is dependent upon the uses to which the data are to be put and so cannot be established for worldwide use." [1]

Household "The concept of household is based on the arrangements made by persons, individually or in groups, for providing themselves with food or other essentials for living. A household may be either: a) a one-person household, that is to say, a person who makes provision for his or her food or other essentials for living without combining with any other person to form part of a multi-person household; or b) a multi-person household, that is to say, a group of two or more persons living together who make common provision for food or other essentials for living... those households without a shelter that would fall within the scope of living quarters. They carry their few possessions with them, sleeping in the streets, in doorways or on piers, or in any other space, on a more or less random basis.” [2]

Housing Unit "A housing unit is a separate and independent place of abode intended for habitation by a single household, though it may be occupied by more than one households..." [3]

Different definitions may be more or less appropriate in different circumstances. For example, one can define a household as "those who share a cooking pot," and that definition may work better in a refugee camp than the definitions given above. For almost all surveys of a general population, rules for defining group quarters must be created, and rules for a term of residency, for example, those who have shared a cooking pot for the past four weeks, must also be created. Most important, because the structure of a family or household is culturally-dependent, local cultural norms must be considered when a definition for a family or household is created.

1. See Households and Families.

2. See United Nations Demographic Yearbook Review.

3. See Consumption and Production Patterns.

Field Test
Field Test

field test

pre-test

The size of the sample for the field test, and the number of days spent on the field test, will depend on issues such as funding, staff availability, and whether problems with the survey procedures or the questionnaire are found. The field test may occur in several stages or as one step. It may be days, or even years, before the actual fieldwork begins, depending on the size of the survey and the extent of the questionnaire-design effort.

The data from the field test are usually not included in the final data for the survey.

Focus Group
Focus Group [1]

focus group

focus group discussion

In the social sciences and urban planning, focus groups allow interviewers to study people in a more natural setting than a one-to-one interview. In combination with participant observation, they can be used for gaining access to various cultural and social groups, selecting sites to study, sampling those sites, and raising unexpected issues for exploration. Focus groups have a high apparent validity, since the idea is easy to understand, the results are believable. Also, they are low in cost, one can get results relatively quickly, and they can increase the sample size of a study, since the interviewer is able to talk with several people at once.

However, focus groups also have disadvantages: the researcher has less control over a group than a one-on-one interview, and thus time can be lost on issues irrelevant to the topic; the data are tough to analyse because the talking is in reaction to the comments of other group members; interviewers need to be highly trained; and groups are quite variable and can be difficult to gather together.

1. This definition extracted from the Wikipedia definition for focus groups (accessed 28 December 2006) [disclaimer].

Frame (Sample Frame)
Sample Frame

sample frame

sample

sample frame

random sample

When a list of the population of interest is not available, an alternate method for capturing the population must be found. Most surveys carried out by governmental statistical agencies rely on a sample frame that is composed of maps that partition the entire country into enumeration areas. In that case, a multi-stage sample design is required. Enumeration areas are first randomly sampled, and then individual housing units are sampled from within the enumeration areas. Finally, individuals are sampled from within the housing units.

Even though the set of maps of enumeration areas is not a list of individuals in the population, it is still considered a sample frame. In that case, however, it is a sample frame of individuals that reside in housing units, not of the total population. Any individual who does not live in a housing unit, for example, a homeless person, is not covered by the sample frame.

Fundar
Fundar, Centre for Analysis and Research

democracy

public policies

Fundar was created in January 1999 by a group of leaders from different disciplines, with the objective of developing schemes for citizen participation, identifying models of action that have been successful in other countries, and experimenting with new methodologies that can help resolve specific social problems.

The organisation's guiding principles are to widen and strengthen citizen participation; demand transparency and accountability from government institutions; contribute to the fulfillment of the rule of law; promote substantive equality; and work for the respect of human rights.

Over the past few years, Fundar has concentrated its work on several areas:

Transparency, public spending and public policies, with projects such as the Latin American Index of Budget Transparency, which is an annual index evaluating the transparency of ten countries from the region in order to issue policy recommendations to their governments. There is also a series of similar projects focusing on health, education and gender, etc.
Human rights commissions and public security, which includes the Metagora activity and various other projects, such as the Citizen Monitoring of Public Human Rights Commissions.
Monitoring of and dialogue with the legislative branch which involves monitoring and systematising information related to the daily activities of two Committees, the Budget and Public Accounts and the Gender and Equity Committees, and communicating with Congressional offices.
Strengthening local capacities and citizen-monitoring mechanisms, with the aim of increasing and strengthening citizen participation in States and/or municipalities and their capacity to influence States’ human rights agendas, by supporting the technical and institutional development of civil society organisations.

Federal District of Mexico

For further information on this pilot survey, please turn to the Synthesis Report produced by the Metagora Coordination Team.

Global Campaign on Urban Governance, The
The Global Campaign on Urban Governance [1]

UN-HABITAT

poverty

urban governance

1. Taken from www.unhabitat.org (accessed 28 December 2006).

Governance / Good Governance
Governance / Good Governance [1]

But what exactly is meant by these two terms?

According to the UN paper, What is Good Governance?, the term “governance” means “the process of decision-making and the process by which decisions are implemented (or not implemented)” [2]. In other words, dealing with governance involves the analysis of the processes and systems by which a specific society, or organisation, operates. Though government is one of the main actors of governance, it is far from being the only one; depending on the specific entity under study, other actors can include “influential land lords, associations of peasant farmers, cooperatives, NGOs, research institutes, religious leaders, finance institutions, political parties, the military, [… as well as the] media, lobbyists, international donors, multi-national corporations, etc.” [3]. Moreover, governance applies to several contexts: corporate governance, international governance, and national, regional or local governance.

Good governance is a form of governance that embodies eight specific characteristics, and can be seen as an ideal of governance. Good governance embodies processes that are “participatory, consensus oriented, accountable, transparent, responsive, effective and efficient, equitable and inclusive, and [which follow] the rule of law” [4]. See Table 1 for further information. Moreover, good governance “assures that corruption is minimised, the views of minorities are taken into account, and that the voices of the most vulnerable in society are heard in decision-making. It is also responsive to the present and future needs of society” [5]. Clearly, there is a close relation between good governance and respect for human rights.

Table 1. The eight characteristics of good governance as defined by the United Nations

Source: United Nations paper, What is Good Governance?, www.unescap.org/huset/gg/governance.htm (24 December 2006).

CHARACTERISTIC DEFINITION

Accountability Accountability is a key requirement of good governance. Not only government institutions but also the private sector and civil society organisations must be accountable to the public and to their institutional stakeholders. Who is accountable to whom varies, depending on whether decisions or actions taken are internal or external to an organisation or institution. In general, an organisation or an institution is accountable to those who will be affected by its decisions or actions. Accountability cannot be enforced without transparency and the rule of law.

Consensus-oriented There are several actors and as many viewpoints in a given society. Good governance requires mediation of the different interests in society to reach a broad consensus on what is in the best interest of the whole community and how this can be achieved. It also requires a broad and long-term perspective on what is needed for sustainable human development and how to achieve such development. This can only result from an understanding of the historical, cultural and social contexts of a given society or community.

Effectiveness and efficiency Good governance means that processes and institutions produce results that meet the needs of society while making the best use of the resources at their disposal. The concept of efficiency in the context of good governance also covers the sustainable use of natural resources and the protection of the environment.

Equity and inclusiveness A society’s well being depends on ensuring that all its members feel that they have a stake in it and do not feel excluded from the mainstream of society. This requires that all groups, but particularly the most vulnerable, have opportunities to improve or maintain their well being.

Participation Participation by both men and women is a cornerstone of good governance. Participation could be either direct or through legitimate intermediate institutions or representatives. Representative democracy does not necessarily mean that the concerns of the most vulnerable in society would be taken into consideration in decision-making. Participation needs to be informed and organised, which requires freedom of association and expression and an organised civil society.

Responsiveness Good governance requires that institutions and processes try to serve all stakeholders within a reasonable timeframe.

Rule of law Good governance requires fair legal frameworks that are enforced impartially. It also requires full protection of human rights, particularly those of minorities. Impartial enforcement of laws requires an independent judiciary and an impartial and incorruptible police force.

Transparency Transparency means that decisions made and their enforcement are achieved in a manner that follows rules and regulations. It also means that information is freely available and directly accessible to those who will be affected by such decisions and their enforcement. It also means that enough information is provided and that it is provided in easily understandable forms and media.

1. This definition is essentially based on the UN paper What is Good Governance? www.unescap.org/huset/gg/governance.htm (24 December 2006). Please refer to this article for more detailed information.

2. As defined by the UN in What is Good Governance? www.unescap.org/huset/gg/governance.htm.

3. Ibid.

4. Ibid.

5. Ibid.

Histogram
Histogram [1]

Example: Age of Lung Cancer Patients in British Columbia, Canada.

Source: http://www.bccancer.bc.ca/ (27 December 2006).

A histogram is a graphical display of data like the one shown above. It is composed of rectangles whose width indicates a range of values and whose height represents the number of data points contained within that range (or the percentage of data points contained within that range). For example, in the figure above, each rectangle's width represents a five-year age range (i.e., 0-5 years, 6-10 years, 11-15 years, and so on) and the height of a particular rectangle represents the number of cases of lung cancer diagnosed in people in British Columbia, Canada, whose ages fall into that age range. Each rectangle is called a "bin," because one can think of them as containers that accumulate data and "fill up" at a rate equal to the count of members of that bin.

The histogram provides a graphical summary of the shape of the data's distribution, whether the data are skewed and whether they have one or more modes. It often is used in combination with other statistical summaries, such as the boxplot, which conveys the median, quartiles, and range of the data.

The shape of the histogram is sometimes particularly sensitive to the number of bins. If the bins are too wide, important information might get omitted. For example, the data may be bimodal but this characteristic may not be evident if the bins are too wide. On the other hand, if the bins are too narrow, what may appear to be meaningful information really may be due to random variations that appear because of the small number of data points in a bin. To determine whether the bin width is set to an appropriate size, different bin widths should be used and the results compared to determine the sensitivity of the histogram shape with respect to bin size. Bin widths are usually selected so that there are between 5 and 20 groups of data, but the appropriate number depends on the situation.

An age-sex pyramid is a special form of histogram.

1. This definition is based on The Histogram and Histograms: Construction, Analysis and Understanding.

Household
See 'Family' Above.

Housing Unit
See 'Family' Above.

HRDAG
Human Rights Data Analysis Group (HRDAG)

Official truth commissions in Haiti, South Africa, Guatemala, Peru, Ghana, Sierra Leone, and Timor-Leste;
The International Criminal Tribunal for the Former Yugoslavia;
Non-governmental human rights groups in Cambodia, Guatemala, and Sri Lanka; and
United Nations missions in Timor-Leste and Guatemala.

HRDAG provides assistance with computer networking, backup, and security, as well as building database and classification systems, and with advanced statistical analysis of mass atrocities.

Further information on HRDAG can be obtained as follows:

1. The main web site for HRDAG is at http://www.hrdag.org/.

2. HRDAG was launched within the Science and Human Rights Program of the American Association for the Advancement of Science. It later moved to Benetech, where it is currently housed.

Human Rights
Human Rights

The term came into wide use after World War II, replacing the earlier phrase "natural rights." Human rights have their origin essentially in the US Bill of Rights and the French Declaration of the Rights of Man and the Citizen of the 18th century. Human rights, as they are known today, are based on the Universal Declaration of Human Rights of 1948 and on the many treaties and agreements that have been developed and concluded in the framework of the United Nations. Though human rights were not given great importance in international relations until the 20th century, this situation has changed dramatically. For centuries, international relations were dominated by the notions of sovereignty and domestic affairs, that is, the common and widespread belief was that each State was free to decide, alone, whether it should grant its citizens certain rights or not. However, in the 20th and 21st centuries, human rights have been given increasing importance in the international arena. For instance, human rights are now considered an essential element in the fields of extradition, development cooperation, and international trade [2].

The concept of human rights acknowledges the fact that every single human being is entitled to enjoy his/her human rights without distinction as to race, colour, gender, language, religion, political or other opinion, national or social origin, property, birth or other status. Human rights are conceived of as universal, i.e., they apply equally and without discrimination to all human beings everywhere; inalienable, in that they cannot be taken away from anyone, except in specific situations, such as lawful sanctions (for example, when a person is sentenced to prison, his/her right to liberty may be temporarily restricted); indivisible and interdependent, in that it is necessary to respect and protect all rights equally because rights are not only considered as having the same importance, but the violation of one right can affect the protection of others; and, finally, as fundamental, that is, they do not refer to all aspects of human life but rather to those essential or basic human needs.

As mentioned above, human rights are based on international human rights law, and encompass different categories of entitlements that range from basic guarantees, to freedoms, to a number of economic, social and cultural rights and have been defined more completely as follows: “international human rights are legal entitlements of individuals against the state or state-like entities guaranteed by international law for the purpose of protecting fundamental needs of the human person and his/her dignity in times of peace and war” [3].

Human rights as a concept, is the result of a historical process. In the course of that evolution, human rights have been classified in terms of three generations. As noted by the authors of The Face of Human Rights, “the history of human rights is reflected in their structure. The rights that have found their way into the Universal Declaration of Human Rights and human rights conventions can be divided into several generations, according to their time of origin” [4].

Human rights are commonly divided into three generations:

The first generation refers to civil and political rights. Such rights are essentially designed to protect the individual against state interference, and are immediately applicable. They include guarantees linked to the protection of life (such as the right to life, and the prohibition of torture and inhuman treatment or punishment); the protection against discrimination; procedural guarantees (including guarantees of fair trial and the right to a defence); freedoms (such as freedom of movement, and freedom of opinion); and political rights (including the right to vote and to be elected).
The second generation comprises economic, social and cultural rights. These rights appeared at a slightly later stage as a response to the gradual impoverishment of individuals and class struggles of the 19th century. These rights, the fulfillment of which requires active measures, in the form of specific legislation, policies and programmes, by States, include economic rights, such as the right to work, and the protection of property; social rights, such as the right to health and social security; and cultural rights, which include the right to education and to take part in cultural life.
The third generation of rights refers to collective rights (also known as solidarity rights). These rights, which were conceived in the second half of the 20th century, cover the right to development, peace, and a clean and healthy environment. However, and with the exception of the African Charter on Human and Peoples’ Rights (1981), such rights have not yet been incorporated into any human rights treaty.

1. See the online dictionary Wordreference at www.wordreference.com (24 December 2006)

2. For further information, see Kälin, W., Müller, L., and Wyttenbach, J., The Face of Human Rights, Lars Müller Publishers, Switzerland, 2004, p. 15.

3. Op. cit., p.17

4. Op. cit., p.21.

5. McChesney, A., Promoting and Defending Economic, Social and Cultural rights, AAAS/HURIDOCS, Washington, DC, 2000, p. 6.

Human Rights Data Analysis Group
See 'HRDAG' Above.

Human Sciences Research Council
Human Sciences Research Council (HSRC)

non-governmental organisations

Over the last couple of years, the HSRC underwent major restructuring, aligning its research activities and structures to South Africa's national development priorities, notably poverty reduction through economic development, skills enhancement, job creation, the elimination of discrimination and inequalities, and effective service delivery.

The HSRC also seeks to contribute to the research and development strategy of the HSRC's parent Department of Science and Technology, especially through its mission to focus on the contribution of science and technology to addressing poverty. With its new structures and greatly extended research personnel, the HSRC is well equipped to respond flexibly and comprehensively to these current and emerging needs.

The organisation has equipped itself to respond flexibly and comprehensively to national requirements by focusing its research capabilities on the following interdisciplinary, problem-orientated, research programmes:

Child, Youth, Family and Social Development
Democracy and Governance
Education, Science and Skills Development
Social Aspects of HIV and AIDS and Health
Society, Culture and Identity
Urban, Rural and Economic Development

The HSRC recognises that if its research is to make a significant difference to aspects of national life, it must be responsive to the needs of users and relevant to topical social issues, while at the same time advancing a theoretically informed understanding of society in all its aspects.

Within the framework of Metagora, the HSRC implemented a pilot survey focusing on the realisation of democracy and human rights in the context of South Africa's land reform process. Measuring respect for human rights and effectiveness of democratic process is of particular significance in South Africa, because throughout the century prior to the first non-racial democratic elections in 1994, the Apartheid State emphatically negated these principles regarding the majority black population. The institutions and policies of post-Apartheid South Africa have therefore largely been directed by the imperative to deepen the non-racial system of governance and democracy, and established a human rights culture. Since land ownership had historically been a source of conflict and contention, land is an important issue to examine.

For further information regarding this pilot survey, please turn to the Synthesis Report produced by the Metagora Coordination Team.

Hypothesis Testing
Hypothesis Testing [1]

inferential statistics

In each problem considered, the question of interest is simplified into two competing claims/hypotheses:

the null hypothesis, denoted H₀, and
the alternative hypothesis, denoted H₁.

These two competing claims/hypotheses are not, however, treated on an equal basis; special consideration is given to the null hypothesis. The null hypothesis can be considered the "existing truth," or that which is to be disproved. There must be a great deal of evidence against the null hypothesis if the alternative hypothesis is to be accepted. This structure is analogous to the "innocent until proven guilty" tenet of the United States court system: innocence is the null hypothesis, and the alternate hypothesis of guilt will only be accepted if there is overwhelming proof.

In the vaccine example we gave above, the burden of proof is on the new vaccine. The null hypothesis is that there is no difference in efficacy between the two vaccines. The alternative hypothesis is that the new vaccine is better. In the income-level example, if we are attempting to prove a disparity in income, then the null hypothesis is that the two socio-demographic groups have the same mean level of income, and the alternative hypothesis is that there is a significant difference in mean income levels between the two groups.

In both of these examples, and in all hypothesis testing, we form statistics in order to test our hypotheses. For example, in the case of the vaccines, we need a measurement that indicates the efficacy of each vaccine. Such a measurement is most likely found through a randomised experiment, where a randomly selected half of a pool of test subjects is given the old vaccine, and the other half of the test pool is given the new vaccine. After a set period of time, the number of occurrences of the disease in each pool is measured. The difference between the count of occurrences of the disease for the old vaccine and the count of occurences of the disease for the new vaccine is calculated. If it is sufficiently large, the null hypothesis - that there is no difference between in efficacy between the two vaccines - is rejected. If the difference is not sufficiently large, we fail to reject the null hypothesis.

In all hypothesis testing, the final conclusion once the test has been carried out is always given in terms of the null hypothesis. We either "reject H₀ in favour of H₁" or "do not reject H₀"; we never conclude "reject H₁", or even "accept H₁". Again, this is because the burden of proof rests with the alternative hypothesis, not the null. If we conclude "do not reject H₀", this does not necessarily mean that the null hypothesis is true, it only suggests that there is not sufficient evidence against H₀ in favour of H₁; rejecting the null hypothesis then, suggests that the alternative hypothesis may be true.

1. This definition based on www.cas.lancs.ac.uk/glossary_v1.1/hyptest.html and www.psychstat.missouristate.edu/introbook/ (accessed 28 December 2006).

Ill-treatment
Ill-treatment [1]

Torture

non-governmental organisations

Torture

The main international texts that refer to ill-treatment and torture are:

Ill-treatment is legally referred to as cruel treatment, and inhuman or degrading treatment or punishment, causing varying degrees of suffering less severe than in the case of torture. There is general agreement that torture is characterised by and distinguished from other forms of ill-treatment by the degree of suffering involved, and that forms of ill-treatment other than torture do not have to be inflicted for a specific purpose, though there does have to be an intent to expose individuals to the conditions that amount to, or result in, ill-treatment.

The essential elements constituting ill-treatment that does not amount to torture are [2]: intentional exposure to significant mental or physical pain or suffering; and by or with the consent or acquiescence of the state authorities.

International law also makes clear that all acts of ill-treatment refer to the actions and violations committed by state agents, not by non-state agents, and that lawful sanctions do not constitute ill-treatment in any form, as long as they do not include the methods and acts referred to as constituting ill-treatment.

In order for international bodies to distinguish among the different forms of ill-treatment and assess the degree of suffering involved, attention is focused on the circumstances of each case and the particular victim. This makes it difficult to identify the exact boundaries between the different forms of ill-treatment, because the circumstances and victims will vary. However, it makes the law more flexible because it can adapt to each case.

All forms of ill-treatment are prohibited under international law. This means that even when treatment is not considered severe enough, in legal terms, to amount to torture, the state may still be found to have violated the prohibition on ill-treatment. In any case, ill-treatment is unconditionally prohibited, even during emergencies or armed conflicts.

Forms of ill-treatment that have been found to amount to torture, either alone or in combination with other forms of treatment, include falaka/falanga (beatings on the soles of the feet), severe forms of beatings, electric shocks, rape, and mock executions [3]. There are also many “grey areas” that do not clearly amount to torture, though are clearly acts of ill-treatment, or about which there is still disagreement. Examples include corporal punishment imposed as a judicial penalty, solitary confinement, certain aspects of poor prison conditions, and disappearances. Examples of ill-treatment not amounting to torture were studied in the Metagora activity carried out in Mexico, and include such acts as law enforcement personnel threatening to harm family and relatives, some forms of hitting and insulting, and illegal detention and lack of communication with the outside world.

The main difference between torture and ill-treatment is technical and legal, based on the severity of the violation and the motives for each violation, which are not always easily defined [4].

1. This definition is based on Giffard, C., The Torture Reporting Handbook: How to Document and Respond to Allegations of Torture within the International System for the Protection of Human Rights, The Human Rights Centre, University of Essex, 2000, www.essex.ac.uk/torturehandbook/english.htm (24 December 2006). Please refer to this article for more detailed information.

2. Ibid, section 3.3.2.

3. Ibid, section 3.3.3.2.

4. Further recommended reading: The Protocol of Istanbul.

Impact Evaluation
Impact Evaluation

Impact evaluations are technical exercises that rely on econometric and statistical models. The three main kinds of impact evaluation designs can be identified as: experimental, quasi-experimental and non-experimental with which are respectively associated control groups, comparison groups, and non-participants. Though all three methods provide valid data about the relative effectiveness of a policy compared with other possible interventions, or with doing nothing at all, experimental designs are seen as the most valid and reliable, and are used most often, when feasible.

Experimental designs

As explained by Phil Davies [1], the “purest form of experimental method is the randomised controlled trial (RCT).” The aim of RCT is to separate possible factors influencing an outcome from the policy itself, by constructing two groups of people on the basis of a purely random selection, and exposing them to exactly the same factors, except from the policy under evaluation. “Randomisation does not mean that [both groups] will be identical, but it reduces the influence of extraneous factors by ensuring that the only difference between the two groups will be those that arise by chance" [2].

The main advantage of such a method is the simplicity in interpreting the results. However, this method is not exempt from a number of problems [3]. For instance, as noted by the World Bank, it may be unethical to carry out such an evaluation; indeed, randomisation may be unethical due to the fact that it may deny benefits or services to otherwise eligible members of the population for purposes of the study. Moreover, it may be difficult to ensure that the selection of both groups is totally random; and experimental designs tend to be expensive and time-consuming.

Quasi-experimental designs

Quasi-experimental designs are based on the same logic and objective as experimental designs, that is, exposing two different groups to exactly the same factors, except the policy under evaluation, in order to assess the true impact of the policy, but, in constructing these two groups, use methods other than randomisation.

These alternative methods include matching or reflexive comparisons [4]. Matching techniques involve building a counterfactual by identifying individuals who do not partake in the policy under study but whose essential characteristics are similar to that of policy participants. Though usually quicker and cheaper to implement than experimental designs, matching tends to reduce the reliability of results, because of selection bias, and increase the difficulty of analysing results.

The reflexive comparison involves constructing a counterfactual based on the characteristics of individuals prior to their involvement in the policy under study. Participants are thus compared to themselves before and after their involvement. The main advantage of reflexive methods is that they make possible the evaluation of policies that cover the entire population, not just subgroups. A major limit, however, is that the changes in the situation of a group before and after the implementation of a policy may be linked to a whole range of factors independent from the policy itself.

Non-experimental designs

This alternative method of impact evaluation should be used when a counterfactual group cannot be constructed based on a random selection of individuals, when it is not possible to identify a group of individuals who are not participants of a policy but share essential characteristics with participants, or when a group cannot be identified for "before and after" comparisons.

In non-experimental designs, statistical methods and econometric techniques are used to compare participants and non-participants to a given policy. Such methods take into account the differences between the two groups, and issues such as selection bias, thus allowing the true impact of policy to be measured. As is the case with quasi-experimental designs, this method tends to be cheaper and easier to implement than the experimental method, since it relies on existing data sources. However, the reliability of results is more fragile and, in statistical terms, such a method can involve a series of complex operations.

1. See online version of: Government Chief Social Researcher’s Office, Prime Minister’s Strategy Unit, Guidance Notes for Policy Evaluation and Analysis, Chapter 1: What is Policy Evaluation?, in The Magenta Book, Cabinet Office, London, 2004, p. 7.

2. Ibid.

3. For more information, see the website of the World Bank, section Evaluation Designs.

4. Ibid.

Incentives
Incentives [1]

incentive

Many researchers have examined the effect of providing a variety of non-monetary incentives to subjects. These include token gifts, such as small packages of coffee, ball-point pens, postage stamps, key rings, trading stamps, participation in a raffle or lottery, or a donation to a charity in the respondent's name. Generally, but not consistently, non-monetary incentives have resulted in an increased response. A meta-analysis of 38 studies that used some form of an incentive revealed that monetary and non-monetary incentives were effective only when enclosed with the survey. The promise of an incentive for a returned questionnaire was not effective in increasing response. The average increase in response rate for monetary and non-monetary incentives was 19.1 percent and 7.9 percent, respectively.

Most researchers have found that higher monetary incentives generally work better than smaller ones. One researcher proposed a diminishing return model, where increasing the amount of the incentive would have a decreasing effect on response rate. A meta-analysis of 15 studies showed that an incentive of 25¢ increased the response rate by an average of 16 percent, and $1 increased the response by 31 percent.

A concern over the use of incentives is that the survey results may be positively or negatively biased because respondents are given an incentive to participate. Several studies have indicated that the use of incentives reduces to some extent item non-response and "bad answers," such as "don't know" or "no answer." It was also noted in a study published by Public Opinion Quarterly that respondents who received incentives have lengthier answers to open-ended questions. Though it seems likely that offering an incentive would bring apathetic participants to the study, research has proven otherwise. The data quality with an incentive, therefore, can actually be considered higher than if the incentive was not offered, as respondents have put more thought into answering the survey questions. There is also evidence that providing incentives will increase respondents' willingness to participate in future studies because they complete the survey feeling positive about the overall experience.

There is also the question of demographic bias. Is a certain group of people more likely to respond to a particular incentive, thus biasing the results toward that demographic group? For example, is it reasonable to assume that offering a monetary incentive would cause greater response from a low-income demographic group? It also makes sense that offering a beauty product would generate biased results because more women would respond than men. A research study conducted at Penn State University indicates, however, that the kind of incentive offered does not affect the response rate differently for different demographic groups. When presented with an incentive, people generally feel obligated to return the favour regardless of the type of gift they received. There is no indication of a group-specific effect of incentives.

There are situations, however, in which providing incentives to respondents may not be worth the additional response generated by the incentives. In the case of surveys about human rights abuses, the subsequent data may be used in a court of law to try perpetrators accused of crimes against humanity. If so, then defence lawyers might argue that respondents had been "paid" to respond in such a way as to put their client in a negative light. When considering the use of incentives, one must carefully weigh the benefits against the potential drawbacks.

1. This definition is comprised, in part, of subsets of the following web sites: www.nbrii.com/Customer_Surveys/ and www.statpac.com/surveys/incentives.htm (accessed 27 December 2006).

Independence
Independence

neutral professionalism of the persons engaged in collecting data and elaborating results and interpretations;
correct technical conception of survey questionnaires, samples an processing;
protecting sources against threats or punishment for their participating to the surveys;
equally, protecting the interviewers and staff against similar threats; and
neutral dissemination of the findings.

Independent Panel of Experts of Metagora
See 'IPE' Below.

Indicator
Indicators [1]

indicator

indicator

An indicator does not have to come in numeric form. One example is the Freedom House “Freedom in the World” Indicator, which classifies countries as free, partly free or not free.

There are a plethora of indicators on governance that are used by governments, development agencies, non-governmental organisations, the media, academic institutions and the private sector. The indicators are often intended to inform users on business investment, allocation of public funds, civil society advocacy or academic research. From a development perspective, governance indicators can be used for monitoring and evaluating governance programmes and projects. Governance indicators are also often used to establish benchmarks, objectives, targets, and goals in the development context.

Numeric indicators come in two basic forms: simple indicators and complex indicators. Simple indicators are usually composed of single numerical values or statistics and reflect relatively simple concepts. Complex indicators are comprise multiple measurements put together in a form that allows the indicator to measure a complex concept or process. For example, the Human Development Index is an overall measurement of human well being that is composed of a weighted average of nation-level statistics on life expectancy at birth, adult literacy rate, school enrollment, and gross domestic product (GDP) per capita.

When choosing an indicator to use, one must check for the following:

validity (does the indicator measure what it purports to measure?);
reliability (can the indicator be produced by different people using the same coding rules and source material?);
measurement bias (are there problems with systematic measurement error?);
lack of transparency in the production of the indicator;
capacity to be representative (for survey data, what is the nature of the sample of individuals?);
variance truncation (the degree to which scales force observations into indistinguishable groupings);
information bias (what kinds of sources of information are being used?), and
aggregation problems (for complex indicators, to what degree are aggregation rules logically inconsistent or overcomplicated?).

1. Adapted from Governance Indicators: A User's Guide (accessed 28 December 2006).

Indigenous Cultural Communities (ICCs)
Indigenous Cultural Communities (ICCs)/Indigenous Peoples (IPs), Communities, or Groups

One of the first organisations to consider the rights of indigenous peoples was the International Labour Organisation (ILO). Indeed, since the 1920s, the ILO has been concerned with indigenous peoples and adopted related conventions in 1957 and 1989. The latter, the Convention No. 169 concerning Indigenous and Tribal Peoples in Independent Countries [1], is considered one of the main international instruments to protect the rights of indigenous peoples.

What are indigenous peoples? Though widely used in international texts, this term has never been formally and precisely defined. Indigenous peoples are the original inhabitants of many countries. The definition most generally used, including by the United Nations, is that elaborated by Jose Martinez-Cobo, the Special Rapporteur to the Sub-commission on Prevention of Discrimination and Protection of Minorities. In his report, which was submitted in the early 1980s and entitled, “A Study of the Problem of Discrimination Against Indigenous Populations,” Martinez-Cobo states that: “Indigenous communities, peoples and nations are those which having a historical continuity with pre-invasion and pre-colonial societies that developed on their territories, consider themselves distinct from other sectors of societies now prevailing in those territories, or parts of them. They form at present non-dominant sectors of society and are determined to preserve, develop, and transmit to future generations their ancestral territories, and their ethnic identity, as the basis of their continued existence as peoples, in accordance with their own cultural patterns, social institutions and legal systems" [2].

The notion of self-identification is essential to the definition of “indigenous groups, communities, or peoples.” Despite some common characteristics, no single definition of “indigenous” exists that reflects their multiplicity, and, in light of such a huge diversity, the self-identification of a person or peoples as indigenous is seen as fundamental to determining indigenous groups. The identification, or ascription, of a person or group as indigenous by others should also be added to the notion of self-identification.

1. See www.unhchr.ch/html/menu3/b/62.htm

2.See the United Nations Concept of Indigenous Peoples.

3. Further information available through the Indigenous Peoples Rights Act of 1997 of the Philippines.

Indigenous Peoples (IPs), Communities, or Groups
See 'Indigenous Cultural Communities (ICCs)' Above.

Indigenous Peoples' Rights Act of the Philippines
See 'IPRA' Below.

Inferential Statistics
See 'Descriptive Statistics' Above.

Interview
Interview [1]

In the case of data collection, an interview is administered for the collection of information. The interviewer usually asks questions as they are outlined on a questionnaire or survey instrument. Such interviews might take place in person or on the telephone.

A focus group is a specific type of interview during which several respondents respond to the interviewer's questions as a group.

1. This definition modified from the Wikipedia definition of an interview (accessed 28 December 2006) [disclaimer].

Interviewee
See 'Respondent' Below.

Interviewer
Interviewer, Statement Taker, Enumerator

interviewer

statement taker

focus group discussions

facilitator

moderator

interview

recorder

census

enumerator

Interviewers must be trained in techniques that will limit interviewer bias. Interviewer bias is the intentional or unintentional prompting by an interviewer that affects the interviewee's response during an interview. Such prompting may lead to distorted results as a consequence of politeness, social acceptability or conflict avoidance [1]. Interviewers should also be trained in methods of persuasion, methods for asking sensitive questions, voice control, body-language control, and confidentiality in order to improve response rates.

1. See www.fbinnovation.de/en/glossary/ (accessed 28 December 2006).

Inventory of Initiatives
Inventory of Initiatives of Metagora

Inventory of Initiatives

democratic governance

A worldwide survey was carried out to identify current and recent measuring work in the field of democratic governance and human rights issues, and to provide, for the first time, an overview of existing expertise around the world. The inventory primarily serves as an information tool on current available resources, with the aim of fostering exchanges among organisations and enhancing the development of assessment and monitoring tools. This inventory is a dynamic database that provides relevant information and networking tools to organisations and individuals who are implementing or planning to implement evidence-based assessment of human rights and democratic governance. The inventory contains information on the scope, aims, methods and outcomes of recent and current initiatives launched on the five continents. It has been made public in the form of an on-line repository designed by professionals as a user-friendly research tool.

The survey revealed the existence of numerous on-going initiatives around the world that were, until now, largely unknown by the human rights and statistical communities. The Inventory therefore not only provides a list of up-to-date efforts to measure democratic governance and human rights, but also gives an interesting picture of local and national needs and priorities in these fields.

The survey covered a broad range of institutions and initiatives that differ in their nature and objectives, themes addressed, geographical scope, and methodological approaches, with the aim of identifying as many field initiatives as possible. Well known institutions working on a global level, such as the University of Essex’s mapping of main international initiatives, and UNDP and Eurostat, which produce the user’s guide on governance indicators, as well as small civil society organisations interested in measuring one single aspect of democratic governance and/or human rights within their local community, responded to the questionnaire. To a large extent, the Inventory helps incorporate statistical analysis into the monitoring of human rights and democratic governance, and promotes the development of innovative and original methods and the proper matching of quantitative and qualitative approaches.

The current version of the Inventory database is based on a first series of responses to a worldwide Metagora survey, and it will be continuously updated and expanded. If your organisation is conducting relevant work in measuring democratic governance and human rights issues, the method you are applying and the results of this work may be of interest to other organisations and individuals around the globe. We therefore invite you to fill in the questionnaire and to send it back to us.

Please note that the description of the initiatives contained in this inventory is the sole responsibility of the implementing organisations and does not necessarily reflect the views of the OECD or of the donor institutions that are supporting the Metagora project.

IPE
Independent Panel of Experts (IPE) of Metagora

Indigenous Peoples Rights Act (IPRA)

PIOs

The IPE is a scientific assessing body, and its main responsibilities are to:

Assess the relevance, quality and reliability of methods and tools generated and used in each Metagora activity;
Review the Metagora Partners Group internal evaluation of the process and outcomes of the whole project;
Organise rigorous scientific peer reviews of the project's results and products; and,
Formulate recommendations on the possible follow-up to the project.

For the purpose of reviewing relevant substantive and technical reports emanating from the project, the IPE was appointed in February 2005 and is headed by Jean-Louis Bodin, former president of the International Statistical Institute (ISI). The IPE is composed of six internationally recognised scientists with knowledge and experience in the fields of statistical science and human rights and governance assessment. The other five members of the IPE are: Milva Ekonomi (former Director General of the National Statistical Institute of Albania), Haishan Fu (former Head of Statistics of the Human Development Report Office of UNDP), Kwaku Twum-Baah (former Director General of the National Statistical Office of Ghana and former Chairman of PARIS21), Carlo Malaguerra, (former Director General of the Swiss Federal Statistical Office), and William Seltzer (Fellow Researcher at Fordham University and former Director of the United Nations Statistical Division).

The IPE conducted a first general assessment of the project and delivered a first intermediate report to the Steering Committee of Donors in May 2005. As a general intermediate assessment, the IPE considered that the project had made remarkable progress in a very short time and that a rich body of important and useful substantive results had been produced by several of the national pilot experiences. Later, in April 2006, the IPE formulated a number of comments and recommendations, based on the analysis and opinion of two experts, Herbert Spirer and William Seltzer. These comments and recommendations still need to be complemented by a more in-depth analysis of the whole technical and narrative documentation produced by the Metagora partners, as well as by a peer-review meeting gathering together the reviewers, the IPE members and the experts involved in the design and implementation of the pilot surveys.

IPRA
Indigenous Peoples' Rights Act (IPRA) of the Philippines

Indigenous Peoples (IPs)

The Metagora activity in the Philippines consists of a small but incisive survey-based study implemented in three northern regions of the country with a high concentration of indigenous peoples. The objective of this pilot project was to develop evidence-based assessment methods and tools combining quantitative and qualitative approaches. The study aimed to measure four aspects of the rights of indigenous peoples to their ancestral domains and lands: the indigenous peoples’ perceptions and awareness of their rights, the enjoyment or violations of these rights, the government measures and customary laws for the realisation of these rights, and the availability of mechanisms for redressing violations or fulfilling rights.

This activity’s relevance for policy is strong, as the national policy at stake is the implementation of the IPRA, entered into force to address the marginalisation and powerlessness of the communities of indigenous peoples, estimated to be one-sixth of the national population of the Philippines. This Act intends to redress a historical injustice against indigenous peoples, whose rights, cultural identity and ancestral lands were alienated by means of application of the feudal jura regalia by the Spanish Crown, by, de facto, their successors, the American colonial government, and then by the Philippine Republic. For the first time, IPRA settles the rights of indigenous peoples and establishes bases for a proactive public policy, including implementing mechanisms and the allocation of funds. IPRA recognises and promotes: the rights of indigenous peoples to ancestral domains and lands; the right to self-governance; economic and social rights; and cultural integrity, including indigenous culture, traditions and institutions.

Item Non-response
See 'Non-Response' Below.

Kindgon, John W.
John W. Kingdon

multiple streams model

problem

policy

political

As stressed by Porter and Hicks, this model aims to explain why some issues and problems become prominent in the policy agenda and are eventually translated into concrete policies while others never do so. Kingdon’s starting point is the "garbage can model" of policy-making developed in 1972 by Cohen, March, and Olsen. This model contradicts the rational approach to decision-making, claiming that policies are not the product of rational actions, because policy actors rarely evaluate many alternatives for action and because they do not compare them systematically.

Kingdon’s model underlines the existence of three distinct, but complementary, processes, or streams, in policy-making. It is the coupling of these streams that allows, at a given time and in a given context, for a particular issue to be turned into a policy. These three streams are [1]:

The stream of problems. The rationale behind this stream is that a given situation has to be identified and explicitly formulated as a problem for it to bear the slightest chance of being transformed into a policy. Indeed, a situation that is not defined as a problem, and for which alternatives are never envisaged or proposed, will never be converted into a policy issue.
The stream of policies. The second stream used to explain how an issue rises or falls on an agenda has to do with the stream of policies. This stream is concerned with the formulation of policy alternatives and proposals. An extremely important aspect of this model is the belief that such proposals and solutions are not initially built to resolve given problems, but rather they float in search of problems to which they can be tied.
The stream of politics. Although they take place independently from the other two streams, political events, such as an impending election or a change in government, can lead a given topic and policy to be included or excluded from the agenda. Indeed, the dynamic and special needs created by a political event may move the agenda around.

As such, no stream is decisive to the overall policy process, though all streams are important. It is precisely when all streams meet and coincide that an issue is transformed from a mere topic and/or problem into a concrete policy. What happens then is a compelling problem is linked to a plausible solution that meets the test of political feasibility [2]. It is not always necessary, however, for all three streams to meet simultaneously for a policy to develop. Indeed, in some cases, partial couplings, the meeting of two of the streams, are sufficient, though the whole policy-making process is made more uncertain. Kingdon argues that policy entrepreneurs play a key role in connecting the streams, and that there are different types of couplings. Indeed, couplings can be more or less “tight” or “loose,” depending on the degree to which streams depend on each other for an issue to develop into a concrete policy [3].

Contrary to the stages model, the multiple streams model does not picture the policy-making process as one that involves steps and stages. Rather, it views the policy process as being the result of the intersection of at least two independent streams. In this model, there is no chronological sequence or priority among the streams. On the contrary, streams act and react according to their own logic, until a window of opportunity is opened and two or more streams coincide and coalesce into a policy.

The major strength of this model is that it recognises that the policy process is fluid and non-linear, and that it involves a vast number of actors and forces.

1. For more information, see Boussaguet, L., Jacquot, S., et Ravinet, P., Dictionnaire des politiques publiques, Presses de la Fondation Nationale des Sciences Politiques, 2004, p.217-225.

2. Op. cit., p.19.

3. For further information, see Lemieux, V., L’étude des politiques publiques : les acteurs et leur pouvoir, 2e édition, Les Presses de l’Université Laval, Canada, 2002.

Land Reform
Land Reform [1]

Throughout history, popular discontent with land-related institutions has been one of the most common factors in provoking revolutionary movements and other social upheavals. To those who work on the land, the private landowner's government-enforced privilege of appropriating a substantial portion -- in some cases half, or even more -- of production without making a commensurate, or indeed any, contribution to production is a rank injustice. Consequently, land reform most often refers to a transfer of ownership by a relatively small number of wealthy, or noble, owners with extensive land holdings, such as plantations, large ranches, or agribusiness plots, to individual or collective ownership by those who work the land. Such transfer of ownership may be with or without consent or compensation. Compensation may vary from token amounts to the full value of the land.

This definition is somewhat complicated by the issue of state-owned collective farms. In various times and places, land reform has encompassed the transfer of land from ownership, even peasant ownership in smallholdings, to government-owned collective farms. It has also, in other times and places, referred to the exact opposite: division of government-owned collective farms into smallholdings. The common characteristic of all land reforms is modification or replacement of existing institutional arrangements governing possession and use of land.

1. This definition is taken from the Wikipedia definition for Land Reform (accessed 28 December 2006) [disclaimer].

Lasswell, Harold Dwight
Harold Dwight Lasswell

Broadly speaking, such stages include [1]:

The identification of policy problems, through demands for action;
Agenda setting, or focusing attention on specific problems (this stage is sometimes merged with the previous one);
The formulation of policy proposals, their initiation and development, by policy-planning organisations, interest groups, and/or the executive or legislative branches of government;
The adoption and legitimation of policies through the political actions of government, interest groups, political parties;
The implementation of policies through bureaucracies, public expenditures, and the activities of executive agencies; and,
The evaluation of a policy’s programmatic implementation and impact.

By breaking the policy process up into different stages, and thus clarifying the stages that are required for a policy to be born and exist, this model combines a series of strengths and limits.

Its major strength is that it reduces the complexity of policy-making to manageable analytical units, facilitating our understanding of it. By separating the process into a series of clear and identifiable steps, it allows us to focus on the distinct procedures and activities necessary to the development of a policy, instead of getting lost in the intricacies of the overall policy process. Indeed, this model does not primarily focus on the actors and institutions involved in policy-making, but rather emphasises the fact that policy-making, as a comprehensive process, cuts across and sometimes even links this variety of actors and institutions (i.e., the executive and the legislative branches of government and the courts, civil society, intergovernmental bodies, etc.). As noted by Porter and Hicks in one of their papers on the process of policy formulation, by shifting attention to the process and its component steps, this model “transcends the boundaries of specific institutions and points to the ways in which individuals and groups interact across them” [2]. Based on this model, entire sections of the study of policies, and particularly public policies, have developed over time, whether with regard to the identification of problems and agenda-setting, or to the implementation and evaluation phases.

1. Robert W. Porter, with Irvin Hicks, Knowledge Utilisation and the Process of Policy Formulation: Toward a Framework for Africa, p. 8.

2. Op. cit., p. 9.

Linear Regression
Linear Regression

Example: Linear Regression of Crime versus Expenditure on Police.

Source: http://napaneedss.limestone.on.ca/greer/mdm4u/ (28 December 2006).

Linear regression is a method for determining the relationship between two or more related numeric variables (population characteristics) within a set of data. It is the inferential statistics method that corresponds to the use of a scatterplot for Exploratory Data Analysis.

An example of a linear regression is given in the figure above. The basic idea is that a line is fit to the data in such a way as to minimise, overall, the distances between the data points and the line. That line then represents the basic relationship between the two population characteristics graphed in the scatterplot - in this case, levels of crime and spending on police.

A statistical test can be done to determine if the slope of that line is significantly different than zero. If it is determined to be positive - that is, the line slopes upward to the right - then the two variables are said to be positively associated. If that slope is determined to be negative - that is, the line slopes downward to the right - then the two variables are said to be negatively associated. If the statistical test does not find the slope of the line to be significantly different than zero, then the two variables are not associated.

More technical information on linear regression can be found at:

www.statcan.ca/english/

http://phoenix.phys.clemson.edu/tutorials/

www.stat.yale.edu/Courses/

Line Plot
Line Plot / Line Graph [1]

Example: Cell Phone use in Anytowne, 1996-2002.

Source: www.statcan.ca/english/ (27 December 2006).

Line graphs are more popular than all other graphs combined because their visual characteristics reveal data trends clearly and they are easy to create. Specifically, a line graph is a set of quantitative (numeric) data, plotted in an x-y coordinate system, that are connected together by lines. The data used to create a line graph must be composed of pairs of observations. For example, a time series is composed of pairs of values: measurements or statistics of some type, and the time periods during which those measurements were taken or for which those statistics were estimated.

An example of a line graph is above. This is a multiple line graph, in which more than one set of data are presented. In this case, two time series are compared: cell phone use by men and cell phone use by women. It is easy to see from this graph that the total cell phone use has been rising steadily since 1996, except for a two-year period (1999 and 2000) where the numbers drop slightly. The pattern of use for women and men seems to be quite similar with very small discrepancies between them.

Advantages of using line graphs include:

They are good at showing specific values of data, meaning that given one variable the other can easily be determined.
They show trends in data clearly, meaning that they visibly show how one variable is affected by the other as it increases or decreases.
They enable the viewer to make predictions about the results of data not yet recorded.

Unfortunately, it is possible to alter the way a line graph appears to make data look a certain way. This is done by either not using consistent scales on the axes, meaning that the value in between each point along the axis may not be the same, or, when comparing two graphs, using different scales for each. It is important to be aware of how graphs can be made to look a certain way, even if that is not what the data are indicating. There will also be data for which a scatterplot is a better solution. A scatterplot is more appropriate when there are multiple values on the y-axis possible for a single value on the x-axis.

1. This definition is based on, and at some points quoted from, www.mste.uiuc.edu/courses/ci330ms/youtsey/lineinfo.html and www.statcan.ca/english/.

Local Consultations (Pre-Questionnaire Design)
See 'Expert Interviews' Above.

MADIO Project
MADIO Project

This project is based on a multi-disciplinary and participatory approach, and was implemented by two scientific partners: the National Institute of Statistics in Madagascar (INSTAT) and the French Institute of Research for Development (IRD, formerly Orstom). The overall project was carried out in close cooperation with researchers from the DIAL.

The MADIO project concentrates on promoting economic analysis in Madagascar, and on rehabilitating the national statistics apparatus. Activities within the MADIO project can be divided into five principal functions:

Implement statistical surveys;
Implement macro-economic models;
Carry out economic studies;
Promote the use and dissemination of results; and,
Participate in training activities.

In all these areas, the MADIO project has been such a success that the project was extended for three years, from 1999 and 2001. The scientific skills it has contributed to develop are many and solid, and its social purpose has been acknowledged by its main partners and users.

The MADIO project is based on an original methodology: a survey in several phases known as the 1-2-3 survey. This 1-2-3 survey methodology consists of three different, but connected, surveys that provide information on the labour market (phase 1), on the informal sector (phase 2), and on levels of household consumption and poverty (phase 3).

For further information regarding this pilot survey, please turn to the Synthesis Report produced by the Metagora Co-ordination Team.

Margin of Error
Margin of Error [1]

Indigenous Cultural Communities/Indigenous Peoples (ICCs/IPs)

confidence interval

A margin of error is usually prepared for one of three different levels of confidence: 99%, 95% and 90%. The 99% level is the most conservative, while the 90% level is the least conservative. The 95% level is the most commonly used.

The margin of error only takes into account random sampling error and the size of the sample. It does not take into account other potential sources of error, such as bias in the questions, bias due to excluding groups that could not be contacted, people refusing to respond or lying, or miscounts and miscalculations.

1. Based on the Wikipedia definition for the margin of error (accessed 28 December 2006) [disclaimer].

Mean
Mean

parameter

statistic

The basic formula for the mean (the statistic) is:

Taken from http://en.wikipedia.org/wiki/Mean

Source: Wikipedia [disclaimer]

where n is the size of the simple random sample used to create the statistic, and x_i is one observation, or one individual's (individual i's) height in the example given above. The mean of an entire population (the parameter mean) uses the same formula; the only difference is that the height of every individual is used (and n then becomes N, the size of the population).

Calculating means becomes more difficult in cases where data are collected through a method other than a simple random sample. For information on weighted means, please see the Wikipedia article (26 December 2006) [disclaimer].

Median
Median [1]

quartile

decile

percentile

The median of a finite list of numbers can be found by arranging all the observations from lowest value to highest value and picking the middle one. There is only one value the median can take.

Medians can be confusing because there may be more than one median for a single set of data. For example if there are an even number of data points, and the two middle values are different from each other, then there is no unique middle value. Notice, however, that at least half the numbers in the list are less than or equal to either of the two middle values, and at least half are greater than or equal to either of the two values, and the same is true of any number between the two middle values. Thus either of the two middle values and all numbers between them are medians. The common convention is to find the middle point between the two numbers and call that the median for the dataset.

Comparing the mean to the median for a set of data can give you an idea how widely the values in your dataset are spread apart. For example, imagine that 10 people are riding on a bus in Redmond, Washington, in the United States of America. The mean income of those riders is $50,000 a year. The median income of those riders is also $50,000 a year.

Now imagine that John Doe gets off the bus and Bill Gates gets on. The median income of those riders remains $50,000 a year. But the mean income is now somewhere in the neighborhood of $50 million or so because Bill Gates' salary is so high.

We now could say that the average income of those bus riders is 50 million dollars. But those other nine riders didn't become millionaires just because Bill Gates got on their bus. The median is therefore a better measure of the "average" bus rider than the mean.

1. This definition is derived from www.robertniles.com/stats/median.shtml and http://en.wikipedia.org/wiki/Median [disclaimer].

Metagora
Metagora

Historically speaking, the potential of statistical analysis for enhancing rigour and reliability in reporting on human rights has been demonstrated by pioneer work undertaken since the 1980s, and by a series of successful projects implemented during the 1990s. Nevertheless, it was only in 2000 that these issues were broadly debated during the International Conference on “Statistics, Development and Human Rights,” held in Montreux, Switzerland. The large, worldwide attendance at this conference, the vigorous political messages delivered by several heads of United Nations agencies, European Commissioners and national ministers, and the unexpected amount of technical and scientific contributions submitted (half of them coming from the developing world), showed the shared interest and strong will of the North and the South to enhance evidence-based assessment of human rights and democratic governance with proper capacities and tools.

The Conference allowed qualified representatives of three international communities – human rights practitioners, statisticians and development experts – to gather together for the first time and share expectations, experiences and views on why and how to develop evidence-based assessment and monitoring. Participants concluded that the issues raised in the Conference merited further exploration. Metagora is an example of a project born out of this logic, aiming to combine statistics and human rights issues.

Metagora, launched as a pilot project in 2004 for a two-year period, is policy-oriented in scope, multi-disciplinary in approach, inclusive and participatory in method. It is based on a North/South partnership. Seven organisations signed Partnership Agreements with the OECD for its implementation. The project therefore operates as a decentralised laboratory for innovation: it is the first international project on measuring human rights and democratic governance to undertake several pilot activities in different regions of the world simultaneously and in an interactive fashion.

The Metagora project aims to:

Identify, document and raise awareness about current and recent work and initiatives in the field of measuring democracy, human rights and governance. The focus is on initiatives undertaken by national and local organisations;
Develop and enhance methods and tools that allow the development of data and indicators upon which beneficiaries can formulate or evaluate policies promoting democracy, human rights and governance;
Provide on the basis of the activities carried out and of the results obtained, outline guidelines on measuring, assessing and monitoring democracy, human rights and governance; and,
Facilitate an informal process among the main producers of governance indicators, aimed at establishing a platform and a road map for a progressive international agreement on key indicators.

The originality of Metagora, in comparison to other existing international initiatives and projects in the field of democratic governance and human rights assessment, lies in its bottom-up approach to the development of internationally agreed indicators and related measurement methods. Metagora partners all work with tools and methods that are designed for a particular issue in a particular local context. However, these are tested to prove their strengths and capacity to produce policy-relevant results, and will thus provide lessons and experiences that can be applied elsewhere in the world.

During the pilot phase, the Metagora activities included:

Three pilot surveys on human rights/governance issues carried out in the Federal District of Mexico (on irregularities, abuses and ill-treatment by law enforcement bodies), in the Philippines (on the rights of indigenous peoples) and in South Africa (on the implementation of land reform);
Two regional multi-country surveys on democracy and governance issues progressively implemented by official statistical agencies in eight capital cities of francophone Africa (since 2001) and in three countries of the Andean Community (since 2002);
The development of indicators and a database on the right to education in Palestine;
The establishment of a controlled vocabulary and harmonised procedures allowing local NGOs to collect and analyse data on human rights violations in Sri Lanka (since 2001);
A worldwide survey on projects measuring human rights and democratic governance; and,
The identification and production of relevant training materials.

Among the significant lessons learned during the two-year pilot phase of Metagora:

Measuring human rights and democratic governance is technically feasible and politically relevant; data on human rights, democracy and governance can be collected and analysed using statistical tools.
On the basis of this information, it is possible to provide indicators that are relevant and useful for political decision and action.
Quantitative data and qualitative information can and should be interrelated to properly inform assessment of human rights and democratic governance.
Official statistical agencies can be involved in the measurement of human rights and democratic governance.
Statistical analysis and quantitative indicators bring a significant value-added to the work of national human rights institutions.
Statistical methods can substantially enhance the research and advocacy of civil society organisations in the fields of human rights and democracy.
Many initiatives in different regions of the world, with approaches and objectives similar to those of Metagora, have been identified and documented.
Through its implementation, Metagora is documenting the experiences, problems encountered and lessons learned in the form of training materials , so these activities (and the tested methods) can be replicated in other countries and contexts.
A North/South network of experts and institutions has been consolidated around Metagora and is continuously growing. This operational network, which is unique in the world, is providing the international community with skills and capacities needed to enhance measuring methods and indicators.

For further information regarding the Metagora project, please turn to the Synthesis Report produced by the Metagora Coordination Team.

Mode
Mode [1]

The mode is not necessarily unique, since the same maximum frequency may be attained at different values. Furthermore, when examining the distribution of numeric data, we may decide that there are several "local" modes. They may have different values than each other, but are modal within a small range of values. If our data have two modes, we call the distribution of the data bimodal. If our data have three modes, they are trimodal. Any distribution with two or more modes can be called multimodal.

The histogram in the following figure displays data that are bimodal. The histogram is composed of life expectancies for each of the countries of Africa and Europe. Although one mode is clearly shorter than the other, each is the tallest within its local "neighbourhood" of data. The pink dot represents the mean of the data, which is shown not to be a good summary description of this bimodal dataset.

Life Expectancy of Females - Africa and Europe.

Source: http://mathrocks.thebernas.net/ (27 December 2006).

Unlike mean and median, the concept of mode also makes sense for categorical data, that is, data that doesn’t consist of numerical values. For example, taking a sample of Korean family names, one might find that "Kim" occurs more often than any other name. Then "Kim" might be called the mode of the sample. However, this use is not common.

1. This definition is based on http://en.wikipedia.org/wiki/Mode_(statistics) [disclaimer].

Multi-stage Sampling
Multi-stage Sampling

The reasons for adopting such a design may be reducing costs, for example, when interviewers are assigned to persons located in a restricted area, or reducing the sample error. Multi-stage sampling is sometimes used when no general sample frame exists. In this case, a first step is to select, at random, a sample of areas, collective units, or villages from a list where they are all registered (primary units). Then, for each selected primary unit, a comprehensive enumeration of all units of lower rank is made, thus obtaining a local sample frame among which a sample of secondary units will be selected.

For example, for each village of the primary sample, a list of all housing units is established, allowing for a selection of a sample of households. Different probabilities can be used at each stage, as well as within one particular stage, for the different units to be selected. Probabilities at the successive stages multiply, so that the resulting probability for selecting one final unit is the product of the probabilities used at each step. The corresponding answers need to be weighted by the inverse of that final probability in order to obtain unbiased estimates.

A cluster sample can be seen as a two-stage sample where the secondary probability is 100 percent.

Multiple Systems Estimation
Multiple Systems Estimation [1]

The basic assumption used to estimate the population size is that the ratio of the number of people captured in both list 1 and list 2 to the number of people captured in list 1 is proportional to the ratio of the number of people captured in list 2 to the number of people in the entire population. In this example, this means:

where N, the total number of people, is unknown. The estimated value of N is then taken as the closest integer to the solution for N in this equality, or

There are many assumptions embedded in this simple ratio solution. Several of these are quite logical, such as the individuals on a list have been randomly sampled from the population, individuals aren’t moving in or out of the population between the creation of the lists, a list never has the same individual listed twice, and the matching across lists is accurate. Another assumption is that there is no dependency between the lists; in other words, the probability that an individual is captured in list 2 is not dependent on whether or not that individual is captured in list 1. A final assumption is homogeneity: that each individual in the population has an equal probability of capture in a given list. If any of these assumptions are violated, capture-recapture may not accurately estimate the population size. If there are additional lists available for the population, however, problems such as list dependency can be addressed through modeling. Using the population from the simple example above, we add a third list in the following Venn diagram:

In this case, we can model for dependencies between the three lists through log-linear models, with constraints, of the following form (where the subscript i [or j or k] for a count is 0 if it does not include people from list i [or j or k], and 1 if it does):

If we believe there are no list dependencies, we use just the first four terms on the right hand side of this equation. If we believe there is only a dependency between the first and third list, we add that interaction term. In this way, a reduced model, one that contains fewer parameters, can be created. Fitting the data from this simple example to the model with a list 1 and 3 interaction term yields an estimated population size of 12.

Other violations of assumptions can also be addressed through modeling or, in certain situations, can be or are cautiously ignored.

1. This definition is adapted from Ball, P., and Asher, J., "Statistics and Slobodan: Using Data Analysis and Statistics in the War Crimes Trial of Former President Milosevic," Chance, 15, No. 4, 2002, pp. 17-24.

Futher information on Multiple Systems Estimation can be found in the following reference: Ball, P., Betts, W., Scheuren, F., Dudukovic, J., and Asher, J., Killings and Refugee Flow in Kosovo, March–June 1999: A Report to the International Criminal Tribunal for Former Yugoslavia. , American Association for the Advancement of Science, Washington, DC, 2002, http://shr.aaas.org/kosovo/icty_report.pdf (26 December 2006).

National Commission on Indigenous People of the Philippines
See 'NCIP' Below.

National Statistical Coordination Board
See 'NSCB' Below.

NCIP
National Commission on Indigenous Peoples (NCIP)

The National Commission on Indigenous Peoples, created in 1997, evolved through a series of government reorganisations in an effort to address the many issues and concerns confronting the country's diverse Indigenous Cultural Communities/Indigenous Peoples (ICCs/IPs), and to deliver basic services to them effectively, efficiently and responsively.

The NCIP works to achieve technically sound and authentic titles, sustainable and culture-sensitive plans, responsive and culture–sensitive programmes and projects, feasible and mission–driven regulations, and expeditious and fair legal services. Whenever possible, it promotes Indigenous Peoples’ consultative mechanisms and bodies at the provincial, regional and national level.

It is composed of three major programmes:

Land Tenure Security, covering cultural mapping of all IP communities, survey and delineation of ancestral domains, and issuance of CADT/CALT;
Establishing Model IP Communities through Development and Peaceo , which includes the development of ancestral domains through the Ancestral Domain Sustainable, Development Protection Plan (ADSDPP), development of people and communities, through delivery of basic services, especially livelihood support, educational assistance, health care, shelter and quick response to address critical situations, and the protection and enhancement of the cultural heritage of Indigenous Peoples; and,
Enforcement and Enhancement of the Human Rights of IPs, referring to adjudication of conflicts through custom laws and tradition and NCIP adjudicatory processes, procedures for the free and prior informed consent of IPs where needed, and legal assistance.

Within the framework of the Metagora pilot project, the NCIP, along with the Commission on Human Rights of the Philippines and the Philippines National Statistical Coordination Board, was involved in implementing a pilot survey focusing on the realisation of the human rights of indigenous peoples. The activity consisted of a small but incisive survey-based study implemented in three northern regions of the country with a high concentration of indigenous peoples. The objective of this pilot experience was to develop evidence-based assessment methods and tools combining quantitative and qualitative approaches. The study aimed to measure four aspects of the rights of indigenous peoples to their ancestral domains and lands: the indigenous peoples’ perceptions and awareness of their rights, the enjoyment or violations of these rights, the government measures and customary laws for the realisation of these rights, and the availability of mechanisms for redressing violations or fulfilling rights.

For further information regarding this pilot survey, please turn to the Synthesis Report produced by the Metagora Coordination Team.

Neyman Allocation
Neyman Allocation [1]

sample design

The precision required for the statistics created from the data collected, and
The resources available to use in the implementation of the survey (i.e., money, staff, expertise, time).

If unlimited resources were available, then the entire population could be sampled and the precision would be maximised. If no resources are available, then no survey can take place. Usually, there are both limited resources available and a required level of accuracy desired, such as a margin of error of 3% or less).

Statisticians have developed formulas to help determine an optimal sample design given these constraints. When a survey uses a stratified sample and has a fixed budget, the ideal sample allocation plan provides the most precision for the least cost. Optimal allocation does just that. Based on optimal allocation, the best sample size for stratum h would be:

n_h = n × [ ( N_h × S_h ) ÷ √( c_h ) ] ÷ [ ∑ ( N_i × S_i ) ÷ √( c_i ) ]

where n_h is the sample size for stratum h, n is total sample size, N_h is the population size for stratum h, S_h is the standard deviation of stratum h, and c_h is the direct cost to sample an individual element from stratum h. The summation ∑ is across all of the strata; i represents one of those strata. Note that c_h does not include indirect costs, such as overhead costs.

The effect of the above equation is to sample more heavily from a stratum when

The cost to sample an element from the stratum is low;
The population size of the stratum is large; and
The variance within the stratum is large.

Neyman Allocation allows optimal allocation when the overall sample size is fixed; this occurs when time or staff are limited. The equation for Neyman Allocation can be derived from the equation for optimal allocation above, by assuming that the direct cost to sample an individual element is equal across strata. Based on Neyman Allocation, the best sample size for stratum h would be:

n_h = n × ( N_h × S_h ) ÷ [ ∑ ( N_i × S_i ) ]

where n_h is the sample size for stratum h, n is total sample size, N_h is the population size for stratum h, and S_h is the standard deviation of stratum h. Again, the summation ∑ is across all of the strata; i represents one of those strata.

1. This definition is based on http://stattrek.com/Lesson6/SampleSizeStrata.aspx (accessed 28 December 2006).

NGO
Non-governmental Organisation (NGO)

governance

Non-governmental Organisation
See 'NGO' Above.

Non-response
Non-response and Item Non-response [1]

Non-response

Item non-response

total non-response

Fundamental Principles of Official Statistics

There are several potential reasons for total non-response. In a household-based survey, the housing unit might be vacant or destroyed. If the housing unit is intact, the occupants might not be at home, or the interviewer might not be able to locate the housing unit. Assuming the interviewer locates the housing unit and at least one occupant is home, that occupant might be unable to answer due to illness, the need to care for children, or another reason. Finally, the occupant might refuse to participate in the survey all together.

Refusals may result from apathy, fear of invasion of privacy, fear of direct threat, from the police or others, or any number of reasons. Some refusals are partial, where the respondent will answer some questions but not all. The end result is item non-response. Others are temporary, where the respondent will answer on the second or third contact.

The consequences of non-response can vary. As non-response increases, the potential for a biased sample increases. This means that the obtained responses of a random sample may no longer be representative of the larger population. In short, response bias can reduce a random sample to what is essentially a convenience sample and, consequently, the conclusions are much weaker.

1. Modified from http://edis.ifas.ufl.edu/PD008 (accessed 29 December 2006).

NSCB
National Statistical Coordination Board (NSCB)

United Nations Statistical Commission

Organisation for Economic Cooperation and Development (OECD)

Within the framework of the Metagora pilot project, the NSCB together with the Human Rights Commission of the Philippines and the National Commission on Indigenous Peoples, was involved in implementing a pilot survey focusing on the realisation of human rights of indigenous peoples. The activity consists of a small but incisive survey-based study implemented in three northern regions of the country with a high concentration of indigenous peoples. The objective of this pilot experience was to develop evidence-based assessment methods and tools combining quantitative and qualitative approaches. The study aimed at measuring four aspects of the rights of indigenous peoples to their ancestral domains and lands: the indigenous peoples’ perceptions and awareness of their rights, the enjoyment or violations of these rights, the government measures and customary laws for the realisation of these rights, and the availability of mechanisms for redressing violations or fulfilling rights.

For further information regarding this pilot survey, please turn to the Synthesis Report produced by the Metagora Coordination Team.

OECD
Organisation for Economic Cooperation and Development (OECD)

democratic government

NGOs

Organisation for Economic Cooperation and Development

The OECD plays a prominent role in fostering good governance in the public service and in corporate activity. It helps governments ensure the responsiveness of key economic areas with sectoral monitoring. By deciphering emerging issues and identifying policies that work, it helps policy-makers adopt strategic orientations. It is well known for its individual country surveys and reviews.

The OECD produces internationally agreed-upon instruments, decisions and recommendations to promote rules of the game in areas where multilateral agreement is necessary for individual countries to make progress in a globalised economy. Sharing the benefits of growth is also crucial, as shown in activities such as emerging economies, sustainable development, territorial economy and aid.

The Metagora project is closely associated with the OECD, since the overall project is coordinated and administered from within the structure of PARIS21, which itself is a consortium created by a series of organisations including the OECD, and hosted by the latter.

For further information regarding the OECD body, structures, and work, please refer to its online page at www.oecd.org/home/.

Organisation for Economic Cooperation and Development
See 'OECD' Above.

Open-ended Question
See 'Closed-ended Question' Above.

Palestinian Central Bureau of Statistics
See 'PCBS' Below.

Panel of Experts of Metagora
See 'IPE' Above.

Parameter
See 'Statistic' Below.

Paris21
Paris 21

Partnership in Statistics for Development in the 21st Century

The PARIS21 consortium is a partnership of policy-makers, analysts, and statisticians from all countries of the world. It focuses on promoting high-quality statistics, making those data meaningful, and designing sound policies. The role of PARIS21 is to foster more effective dialogue among those who produce development statistics and those who use them, through facilitating international events, supporting country-based activities, regional workshops, and subject-matter task teams.

Consortium members are from governments, international organisations, professional bodies, and academic institutions around the world. They have practical experience and a desire to collaborate to improve policymaking through reliable, pertinent statistics. PARIS21 is serviced by a small secretariat hosted by the Development Co-operation Directorate in the Organisation for Economic Cooperation and Development in Paris, France.

The Metagora pilot project is coordinated and administered within the PARIS21 consortium.

For further information, please consult the web site of the consortium at www.paris21.org/.

Partner Implementing Organisation
See 'PIO' Below.

PCBS
The Palestinian Central Bureau of Statistics (PCBS)

www.bom.gov.au/climate/glossary/

Since it was established, the PCBS has been fully engaged in data collection and statistical tabulation in almost 40 fields covering the socio-economic conditions of the Palestinian society, including human rights. Now, after a decade of continued production of statistics from censuses, surveys, and administrative records, the PCBS, as the official central statistical agency for Palestine, is also engaged in capacity building. The Metagora activity with the PCBS is implemented in this context.

The Metagora pilot activity in Palestine examines, through a large participatory process, the possibilities for involvement of official statistical agencies (OSAs), and civil society organisations, in measuring human rights and democratic governance issues. The activities focus on the development of indicators on the right to education based on official statistical data and on information collected by civil society organisations. Quantitative and qualitative information is integrated into a dynamic database developed and managed by the PCBS, which serves as reference tool for independent policy-oriented analyses and reporting by research centres, human rights organisations and political actors.

For further information regarding this pilot activity, please turn to the Synthesis Report produced by the Metagora Coordination Team.

Percentile
Terciles, Quartiles, Quintiles, Deciles, and Percentiles

descriptive statistics

The 1st percentile cuts off the lowest 1% of data.
The 98th percentile cuts off the lowest 98% of data.
The median is the 50th percentile, and cuts off the lowest 50% of the data.

When percentiles are estimated by ranking the items of a finite sample, the percentile generally falls between two of the observed values, and the midway value is often taken. For example, a class off 22 students receives the following scores on an exam:

50, 50, 52, 58, 58, 61, 64, 64, 67, 68, 72, 74, 76, 78, 78, 79, 83, 85, 92, 92, 96, 98

We have listed the scores in order to make determining the percentiles easier. The median, or 50th percentile, is then between the 11th and 12th observation, or between 72 and 74. We therefore use 73 as the median. Similarly, the 10th percentile can be found by noting that the bottom 2 values are the bottom 9% of the data (2 ÷ 22 ≅ 0.09), and the bottom 3 values are the bottom 14% of the data (3 ÷ 22 ≅ 0.14). The 10th percentile must therefore be between the second and third data point. Since the midway value between those two data points is 51, we find the 10th percentile to be 51.

The terms tercile, quartile, quintile and decile refer to the percentiles that divide the distribution of data into 3, 4, 5, or 10 equal parts, respectively. The relationships between terciles, quartiles, quintiles, deciles, and percentiles are outlined in the table below.

tercile quartile quintile decile percentile

1st 10th

1st 2nd 20th

1st 25th

3rd 30th

1st 33rd

2nd 4th 40th

2nd 5th 50th

3rd 6th 60th

2nd 67th

7th 70th

3rd 75th

4th 8th 80th

9th 90th

3rd 4th 5th 10th 100th

Further information on terciles, quartiles, quintiles, deciles, and percentiles can be found in the following resources:

www.answers.com

http://en.wikipedia.org/

disclaimer

Piechart
Piechart [1]

Example: Avenue High School Student and faculty response to the poll 'Should Avenue High School adopt student uniforms?'

Source: www.statcan.ca/english/ (27 December 2006).

A pie chart is a way of summarising a set of categorical data. A pie chart takes the form of a circle divided into a series of segments. Each segment represents a particular category. The area of each segment is the same proportion of a circle as the category is of the total data set; in other words, the segment represents the percentage of the total data set that is composed of the particular category associated with that segment. Often, a segment of the drawing is separated from the rest of the pie in order to emphasise an important piece of information, as in the pie chart given above.

The use of the pie chart is quite popular, as the circle provides a visual concept of the whole (100 percent). Pie charts are also one of the most commonly seen charts because they are simple to use. Despite its popularity, pie charts should be used sparingly for two reasons: first, they are best used for displaying statistical information when there are no more than six components: otherwise, the resulting picture will be too complex to understand. Second, pie charts are not useful when the values of each component are similar because it is then difficult to see the differences between slice sizes.

When drawing a pie chart, ensure that the segments are ordered by size (largest to smallest) in a clockwise direction. Labeling the segments with percentage values often makes it easier to tell quickly which segment is bigger. Whenever possible, the percentage and the category label should be indicated beside their corresponding segments. This way, users do not have to look back at the legend in order to identify what category each colour represents.

When displaying statistical information, refrain from using more than one pie chart for each figure. A user might find it difficult to compare a segment from one pie chart to the corresponding segment of the other pie chart. However, in a split barplot, such as an age-sex pyramid), these segments become bars that are lined up back-to-back, making it much easier to make comparisons.

1. This definition is based on Pie Charts.

PIO
Partner Implementing Organisation (PIO)

Though Metagora is coordinated by a central team based in Paris, its implementation relies on a multi-disciplinary and inclusive community of organisations and individuals. The core of this community is formed by representatives and experts from the seven Partner Implementing Organisations (PIOs) that signed Partnership Agreements with the OECD for implementing Metagora. The project therefore operates as a decentralised laboratory for innovation.

These organisations, which are scientific, government and non-governmental bodies, are:

In the activity implemented in Sri Lanka, the link with Metagora is not based on a formal partnership agreement with the OECD; nevertheless, the Asia Foundation and the member organisations of the Human Rights Accountability Coalition of Sri Lanka are de facto considered and treated by the Metagora community as key project partners.

The main tasks of these organisations are to:

Manage, organise, and report on the progress of local activities;
Contract, supervise and coordinate local experts;
Set participatory local teams and fora;
Prepare for, participate and engage in follow-up to Metagora Partners’ Group activities;
Develop interaction and synergies among the different pilot activities, conduct direct exchanges and carry out visits and assistance missions;
Participate actively in the meetings of the Metagora Forum; and,
Contribute to the global synthesis, to the drawing of global lessons and guidelines and to the finalisation and presentation of the expected Metagora products.

Policy
Policy

policy

policy process

Although definitions are many, a number of key elements are present in all definitions; these relate to actors, activities, problems and solutions. Policy may be defined as "a plan of action to guide decisions and actions oriented towards solving problems. The term may apply to government, private sector organisations and groups, and individuals." A potential definition for public policy is then "a course of action or inaction chosen by public authorities to address a problem, via laws, regulations, and decisions and actions of government" [1].

Policies can take different forms: they can be implemented at various levels (locally, regionally or nationally); they can be aimed at the whole population or at specific subgroups of the population; and some will be essentially regulatory in nature, while others will have a more distributive component. Policies are also dynamic and very often the result of complex power struggles amoung various actors.

1. From Wikipedia (accessed 31 December 2006) [disclaimer].

Policy process
Policy Process

Although there are many definitions, a number of key elements, relating to actors, activities, problems and solutions, are present in all of them.

Policy may be defined as “a plan of action to guide decisions and actions oriented towards solving problems. The term may apply to government, private sector organisations and groups, and individuals.”

Different models have been created in an attempt to simplify and explain the process by which a policy is chosen and carried out. These models include the stages approach, the streams model, the cognitive approach, the socio-historical approach, the institutional model, and the organisational approach.

Given that policy processes are highly context-specific, it is essential to understand the country’s or area’s particular context and characteristics in order to grasp the specificity of its policy-making process, and potentially play a role in shaping it.

Political Rights
Political Rights

Human rights

civil rights

the first generation refers to civil and political rights;
the second generation comprises economic, social and cultural rights; and,
the third generation refers to collective rights.

Political rights, along with civil rights, are primarily designed to protect the individual against state interference, and are immediately applicable. Political rights can be seen as covering the right to political participation, that is, citizens’ right to seek to influence and participate in the public affairs of the society to which they belong. Political participation can take many forms, the most notable of which is included in the right to vote. However, it also covers the right to join a political party; the right to stand as a candidate in an election; the right to participate in a demonstration; and freedom of association. Though political and civil rights are distinct, the difference between the two is not always obvious or clear; indeed, they sometimes overlap. The freedom to express one’s opinion, and the freedom of association, for example, are clearly linked to the right to political participation, and thus are political rights, but they are often also seen as civil rights.

The right to political participation merits special attention, as it is restricted, to a large though not absolute extent [1], , to citizens. Whereas the other rights recognised by the UN Charter, the Universal Declaration of Human Rights, and the International Covenant on Civil and Political Rights inhere in human beings on the basis of their status as human beings, the right to political participation is, in part, limited to people endowed with the status of citizen. Such a status is linked to the context of a political community and, most significantly, a government. The right to political participation therefore presupposes the existence of a government [2].

Though distinct, civil rights and political rights are closely linked; their protection and fulfillment depends to a large extent on that of the other. All human rights are indivisible, interdependent and interrelated, such that the fulfillment and protection of civil and political rights depends on, and influences, other categories of human rights.

In international human rights law, political rights are protected by the International Covenant on Civil and Political Rights (ICCPR). The ICCPR, drafted in 1966, entered into force in 1976, and is monitored by the Human Rights Committee. Over time, additional protocols and instruments have been created that also aim to contribute to the protection of political rights.

All States Parties to the Covenant are required to submit regular reports to the Committee on how they are realising and protecting political and civil rights. Such information is provided through self-reporting and thus might be limited. The reports provided are examined by the Committee, which then disseminates its concerns and recommendations in the form of “concluding observations.” In the framework of the First Optional Protocol to the Covenant, the Committee was given jurisdiction to examine individual complaints; however, this is not yet the case of the Committee set up to monitor economic, social and cultural rights [3].

1. For example, the freedom of opinion is not, as such, reserved to citizens.

2. See Klein, H.,The Right to Political Participation and the Information Society, Georgia Institute of Technology, Atlanta, presented at the Global Democracy Conference in Montreal, 29 May – 1 June 2005.

3. For further information, please refer to the web site of the OHCHR, www.ohchr.org/english/bodies/hrc/index.htm.

Population Pyramid
See 'Age-Sex Pyramid' Above.

Poverty
Poverty [1]

Descriptions of material need, usually including the necessities of daily living, such as food, clothing, shelter, and health care. Poverty in this sense may be understood as the deprivation of essential goods and services.
Descriptions of social relationships and need, including social exclusion, dependency, and the ability to participate in society. This would include education and information.
Describing a lack of sufficient income and wealth. The meaning of "sufficient" varies widely across the different political and economic parts of the world.

When measured, poverty may be absolute or relative.

Absolute poverty refers to a set standard that is consistent over time and between countries. An example of an absolute measurement would be the percentage of the population eating less food than is required to sustain the human body (approximately 2000-2500 kilocalories per day). Another example is The World Bank definition of poverty. The World Bank defines extreme poverty as living on less than US$ (PPP) 1 per day, and moderate poverty as less than $2 a day. Additional indicators of absolute poverty include life expectancy, child mortality, literacy, percentage of children not in the labour force, electric power, cars, radios, and telephones per capita, and the proportion of the population with access to clean water.

Relative poverty views poverty as socially defined and dependent on social context. In this case, the number of people counted as poor could increase while their incomes rise. A relative measurement would be to compare the total wealth of the poorest one-third of the population with the total wealth of richest 1% of the population.

In many developed countries, the official definition of poverty used for statistical purposes is based on relative income. As such, many critics argue that poverty statistics measure inequality rather than material deprivation or hardship. For instance, according to the U.S. Census Bureau, 46 percent of those in "poverty" in the U.S. own their own home, with the average poor person's home having three bedrooms, with one-and-a-half bathrooms, and a garage. Furthermore, the measurements are usually based on a person's yearly income and frequently take no account of total wealth. As another example, the main poverty line used in the OECD and the European Union is based on "economic distance," a level of income set at 50 percent of the median household income.

A relatively new framework for exploring poverty is a rights-based framework, based on economic human rights. According to that framework, poverty is not only lack of income, but it also encompasses deprivations in areas of health, education, participation and security. Human poverty is therefore a denial of human rights because it infringes on human freedom, it destroys human dignity, and it implies discrimination and injustice. In the rights-based framework of poverty reduction, there are claimants of rights and there are duty-holders, which include communities, governments at all levels, private sector, civil society, and external development partners. Poor people therefore have the ability to claim that their right to overcoming poverty has been violated and they can hold the specifically identified duty-holders accountable. Within the rights-based framework, accurate poverty statistics and indicators are essential for enforcing that accountability.

1. This definition is based on http://en.wikipedia.org/wiki/Poverty [disclaimer] and www.undp.org/poverty/docs/employment/HRPR.doc (accessed 27 December 2006).

Principal Components Analysis
Principal Components Analysis [1]

variance

For example, a multi-country survey may collect the following data:

age, sex, marital status, race, ethnicity, ancestry, national citizenship, refugee status, disability status, rural/urban residency, nation of residency, household size, number of dependents, health insurance status, immunisation status, general health status, nourishment status, daily protein intake, vitamin intake, medicines taken, adjusted gross income, alimony received, child support received, employment status, years of employment, worker class (blue collar, white collar, etc.), occupation, education level, school enrollment status, hours worked per week, housing unit type, enrollment in government assistance status, etc.

Interpreting that volume of data may be very difficult. PCA allows those different characteristics of the respondents to be summarised in a few variables, which can then be compared across respondents.

1. This definition is based on the Wikipedia definition for Principal Components Analysis [disclaimer], www.statistics.com/resources/glossary/p/pca.php, and www.spectroscopyeurope.com/TD_16_6.pdf (accessed 28 December 2006).

Probability Sample
See 'Random Sample' Below.

Q-methodology
Q-Methodology [1]

The name "Q" comes from the form of factor analysis that is used to analyse the data. Normal factor analysis, called "R method," involves finding correlations between population characteristics, such as height and age, across a sample of subjects. Q, on the other hand, looks for correlations between subjects across a sample of population characteristics. Q factor analysis reduces the many individual viewpoints of the subjects to a few "factors," which represent shared ways of thinking. It is sometimes said that Q factor analysis is R factor analysis with the data table turned sideways. While helpful as a heuristic for understanding Q, this explanation may be misleading, as most Q methodologists argue that for mathematical reasons no one data matrix would be suitable for analysis with both Q and R.

One salient difference between Q and other social science research methodologies, such as surveys, is that it usually uses many fewer subjects. Indeed, Q is sometimes used with a single subject. In such cases, a person will rank the same set of statements under different conditions of instruction. For example, someone might be given a set of statements about personality traits and then asked to rank them according to how well they describe herself, her ideal self, her father, her mother, etc.

Within the Metagora context, Q-Methodology was used during a pilot survey implemented by the Human Sciences Research Council of South Africa.

1. Based on the definition at http://en.wikipedia.org/wiki/Q_methodology (accessed 28 December 2006) [disclaimer].

Qualitative Data
See 'Data' Above.

Quantitative Data
See 'Data' Above.

Quartile
See 'Percentile' Above.

Questionnaire
Survey Instrument, Questionnaire, Statement

questionnaire

survey instrument

statement form

closed-ended

open-ended questions

The process of administration of the questionnaire may be:

self-administered, such as when the questionnaire is mailed to potential respondents, sent through electronic mail to potential respondents, or made available on a web site;
administered face-to-face by an interviewer or statement taker to one respondent or multiple respondents simultaneously; or
administered over the telephone by an inteviewer or statement taker.

If the list of questions is being administered through a random sample survey, then that list of questions is usually called a questionnaire or survey instrument. If the list of questions is being administered to a set of volunteer respondents, and is composed primarily of open-ended questions, then the list of questions is called a statement or statement form. This is often the case when statements about human rights abuses are collected by non-governmental organisations.

Designing a questionnaire is a complex process. Care must be taken to create questions of clear and simple meaning, laid out in such a way as not to confuse the respondent (in self-administered surveys) or the interviewer (in non-self-administered surveys). In addition:

From a technical point of view, a question must not induce a specific answer. Dishonest researchers or surveyors might draw false evidence from inductive questions to support their thesis. But even sincere researchers might inadvertently lead the respondents towards answers that do not reflect their actual situation or opinion.
From an ethical point of view, a question must avoid “undue stress through participation, loss of self-esteem, psychological injury or other side effects” (International Statistical Institute: Declaration on Professional Ethics §4.4, 1985).

A good questionnaire design process is essential for the creation of a high-quality questionnaire.

Questionnaire Design
Questionnaire Design (Survey Design)

Questionnaire design

Development of the research questions. This is one of the most important steps. If the research questions are well defined, then every question on the questionnaire will directly address one of the research questions.
Initial research, expert interviews. This step is necessary if the questionnaire designer does not have enough information to begin.
Preliminary draft of questionnaire created.
Expert review occurs.
Revision of the questionnaire, based on the comments of the experts. Sometimes, another cycle of the previous step and this step is required if the experts find many problems in the questionnaire.
Translation and back-translation. This is only necessary for surveys that will be administered to respondents who do not speak the primary language in which the questionnaire was written. It is highly recommended that translation occur during questionnaire design, rather than in the field while the survey is being administered.
Revision of the questionnaire in all languages, as necessary. It may be necessary to repeat the previous step and this step until each back-translation matches the original language version.
Testing via cognitive interviewing, in all language versions.
Revision of the questionnaire, based on the results of the cognitive interviewing. The previous step and this step may be repeated for multiple "rounds" of testing.
Language decentering, or simultaneous back-translation of all language versions of the questionnaire, for final comparison across all languages and the initial language. Revision occurs at this step as necessary, with a focus on conceptual equivalence across languages.
Interviewer training. Issues raised by the interviewers as they are trained are noted and the questionnaire is revised, if necessary.
Field testing. Major problems with the questionnaire will probably not be found at this point if all previous steps were taken, but any major problems found may involve returning to a previous step in the questionnaire-design process.
Final revisions.

At that point, the questionnaire is finalised and printed, and the fieldwork for the survey begins.

Quintile
See 'Percentile' Above.

Random Sample
Random versus Non-random Samples [1]

sample

sample size

Samples are selected in such a way as to avoid presenting a biased view of the population. The sample will be unrepresentative of the population if certain members of the population are excluded from any possible sample. For example, if a researcher is interested in the drug-usage patterns among teenagers, but collects the sample from schools, the sample is biased because it excludes teenagers not in school for a variety of reasons, such as lack of funds to attend or schooled at home. Biases may also occur if some members of the population are more likely or less likely to be included in the sample than other members of the population for a reason other than the sample design. So the sample collected from schools is also biased because students who miss a lot of school days because of a chronic illness will be less likely to be selected than students who attend regularly.

The best way to avoid a biased or unrepresentative sample, and thus to obtain a representative sample of the population, is to select a random sample, also known as a probability sample. A random sample is defined as a sample in which every individual member of the population has a non-zero probability of being selected as part of the sample. In a simple random sample, every individual member of the population has the same probability of being selected as every other individual member. Other types of random samples fall under the category of complex sample design.

A sample that is not random is called a non-random sample or a non-probability sample. Some examples of non-random samples are convenience samples, judgment samples, purposive samples, quota samples, and snowball samples.

1. Based on the definition for a statistical sample at Wikipedia (accessed 28 December 2006) [disclaimer].

Range
Range [1]

The range provides an indication of statistical dispersion, or how spread-out the data are. It is measured in the same units as the data. Since it only depends on two of the observations, it is a poor and weak measure of dispersion except when the sample size is large. A better measure of the dispersion of the data is the variance or the standard deviation. The range, however, is much simpler to calculate.

The midrange point, that is, the point halfway between the two extremes, is an indicator of the central tendency of the data, or the "average" value of the data. Again it is not a particularly good measure of central tendency for small samples; other more robust measures of the "average" of the data are the mean and the median.

1. This definition is based on the Wikipedia definition of the range [disclaimer].

RDP of South Africa
Reconstruction and Development Programme (RDP) of South Africa [1]

The ANC's chief aim in developing and implementing the RDP was to address the immense socioeconomic problems brought about by its predecessors under the Apartheid regime. Specifically, it aimed to alleviate poverty and address the massive shortfalls in social services across the country. The RDP attempted to combine measures to boost the economy, such as contained fiscal spending, sustained or lowered taxes, reduction of government debt, and trade liberalisation, with socially minded social-service provisions and infrastructure projects. In this way, the policy incorporated both socialist and neo-liberal elements, but could not be easily categorised in either camp.

The RDP White Paper outlined six principles that would guide and give substance to the remainder of the programme:

The RDP would be an integrated, well-coordinated and sustainable programme, to be conducted in and integrated amoung all three spheres of government, along with civil society, business and parastatals.
The RDP would be people-driven.
The RDP would attempt to play a role in ending the endemic violence within South Africa by embarking on a national drive for peace and security. Such conditions, the Programme noted, would also help to encourage investment, thus fostering economic expansion and greater development.
The commitment of all parties to the RDP would encourage and further the grand project of nation-building.
The RDP would link growth, development, reconstruction, redistribution and reconciliation into a "unified programme", held together by a broad infrastructure programme that would focus on creating and enhancing existing services in the electricity, water, telecommunications, transport, health, education and training sectors.
The success of the first five principles would, in turn, facilitate the sixth: democratisation. The document notes "minority control and privilege" within the economy as a major obstacle to the achievement of an integrated, development-orientated economy. It also acknowledges that the people most affected by economic policy should participate in decision-making, and that the government would also have to be restructured to fit the priorities of the RDP.

Proponents of the RDP argue that the programme oversaw many major advances in dealing with South Africa's most severe social problems:

Housing: Between 1994 and the start of 2001, over 1.1 million cheap houses eligible for government subsidies had been built, accommodating 5 million of the estimated 12.5 million South Africans without proper housing.
Clean water: By the beginning of 1998, standpipes had been installed within 200 metres of the dwellings of about 1.3 million rural people. By August of that year, Minister of Water Affairs Kader Asmal stated that since he had taken office, more than 2.5 million people had been given access to fresh safe water. By 2000, a total of 236 projects had supplied clean piped water to nearly 4.9 million people, most of whom were inhabitants of former homelands.
Electrification: Between 1994 and May 2000 around 1.75 million homes had been connected to the national grid, while the proportion of rural homes with electricity grew from 12 percent to 42 percent.
Land Reform: By 1999 some 39,000 families had been settled on 3,550 square kilometres of land. Authorities claimed that 250,000 people had "received land" within four years.
Healthcare: Between April 1994 and the end of 1998, around 500 new clinics gave an additional five million people access to primary health care. Under the polio-hepatitis vaccination programme that began in 1998, eight million children were immunised within two years.
Public works: A community-based public works programme provided employment over five years to 240,000 people on road-building schemes and the installation of sewage, sanitation facilities and water supplies.

Critics have questioned the scope of change represented by many of the statistics, and have argued that realities on the ground signify a far more modest improvement than the government claims. They have attacked, in particular, the standards of housing and water delivery, healthcare improvements and the success of the land reform policy and agricultural reforms:

Housing: Critics of the RDP cite poor housing quality. One research investigation in 2000 found that only 30 percent of new houses complied with building regulations. Critics also note that new housing schemes are often dreary in their planning and layout, to the extent that they often strongly resemble the bleak building programmes implemented by the Apartheid government during the 1950s and 60s.
Clean water: Critics of the RDP have targeted the government's assertions regarding the provision of clean water, citing an array of problems and complications with RDP policies that have led to their partial or complete failure during the implementation stage.
Land Reform: The number of families settled on land under the RDP was far below the Programme's goal. The RDP had aimed to resettle families on 300,000 square kilometres of land, in reality, only just over one percent of this goal was achieved. In addition, the advances in many other areas of public services came partly through the removal of agricultural subsidies, which resulted in huge job losses. Between 1994 and 1998, the number of workers on commercial farms declined from 1.4 million to just 637,000. Thus, the number of people employed in the agricultural sector actually declined substantially under the RDP.
Healthcare: Critics of the RDP argue that access to healthcare improved only slightly under the RDP and that, even with moderately improved access, standards at many medical institutions declined rapidly. They note that use of healthcare facilities increased by just 1.6 percent between 1995 and 1999, and that even these modest improvements have been eclipsed by the advance of the AIDS pandemic and other health epidemics, such as malaria. Between 1995 and 1998, life expectancy of South Africans fell from 64.1 years to 53.2 years, with AIDS patients sometimes occupying up to 40 percent of beds in public hospitals.

1. This definition is a modification of the Wikipedia definition for the RDP (accessed 28 December 2006) [disclaimer].

Reconstruction and Development Programme of South Africa
See 'RDP of South Africa' Above.

Relational Database
Relational Database

assumptions

Representative Sample
See 'Random Sample' Above.

Respondent
Respondent, Interviewee

respondent

interviewee

In the case of a focus group, the respondents can also be called informants or participants.

Response rate
Response Rate [1]

Response rate

sample

non-response rate

not

sample

The non-response rate should not be confused with the refusal rate. There are many reasons for non-response, only one of which is refusal to participate in the survey.

Methods used to increase the response rate to a survey include the use of incentives, pre-letters, interviewer training on response-encouraging techniques, and follow-up requests.

1. This definition is based on the Wikipedia entries for response rate and statistical surveys (accessed 29 December 2006) [disclaimer].

Sample
See 'Random Sample' Above.

Sample Design
See 'Complex Sample Design' Above.

Sample Frame
See 'Frame (Sample Frame)' Above.

Scatterplot
Scatterplot [1]

Example: Carat Size versus Price of Diamonds on Singapore Market.

Source: www.stat.yale.edu/Courses/ (27 December 2006).

A scatterplot, scatter diagram, or scatter graph is a graph used in statistics to visually display and compare two or more sets of related quantitative (numerical) data.

For example, to study the effects of lung capacity on the ability to hold one's breath, a statistician would choose a group of people to study, and test each one's lung capacity (first variable) and how long that person could hold his/her breath (second variable). The researcher would then set up the data in a scatterplot, assigning "lung capacity" to the horizontal (X) axis, and "time holding breath" to the vertical (Y) axis. A person with a lung capacity of 400 cc who held his/her breath for 21.7 seconds would be represented by a single dot on the scatter plot at the point (400, 21.7) in Cartesian coordinates. The scatter plot of all the people in the study would enable the statistician to obtain a visual comparison of the two sets of data, and help to determine what kind of relationship there might be between them.

A scatterplot is often employed to identify potential associations between two variables, where one may be considered to be an explanatory variable, such as size, in carats, of a diamond, and another may be considered a response variable, such as price of the diamond on the Singapore market. A positive association between size and price would be indicated on a scatterplot by an upward trend (positive slope), where higher prices correspond to higher sizes and lower prices correspond to fewer carats. A negative association would be indicated by the opposite effect (negative slope), where the biggest diamonds would have lower prices than the smallest diamonds. Or, there might not be any notable association, in which case a scatterplot would not indicate any trends whatsoever. In the figure given above, a positive association between carat size of diamond and price of diamond on the Singapore market is shown.

A line of best fit can be drawn in order to study the correlation between the variables. An equation for the line of best fit can be calculated by using the correlation coefficient. That is what occurs during linear regression.

1. This definition is based on http://en.wikipedia.org/wiki/Scatterplot [disclaimer] and http://www.stat.yale.edu/Courses/1997-98/101/scatter.htm.

Simple Indicator
See 'Indicator' Above.

Skew
Skew [1]

histogram

Roughly speaking, a distribution has positive skew (right-skewed) if the right (higher value) tail is longer or fatter and negative skew (left-skewed) if the left (lower value) tail is longer or fatter. The two are often confused, since most of the mass of a right (or left) skewed distribution is to the left (or right) of its respective tail. Examples of right- and left-skewed distributions are given in the figures below. Note that when the histogram is left-skewed, the mean is to the left, thus smaller than the median, and when the histogram is right-skewed, the mean is to the right, thus bigger than the median.

This is a left-skewed distribution.

Source: http://mathrocks.thebernas.net (27 December 2006).

This is a right-skewed distribution.

Source: http://mathrocks.thebernas.net/ (27 December 2006).

Many statistical methods assume data are symmetric about the mean. But in reality, data points are not perfectly symmetric. So, an understanding of the skewness of the data indicates whether deviations from the mean are going to be positive or negative, and whether they are going to be significant enough to undermine the assumptions of any statistical methods used.

1. This definition based on http://www.statcan.ca/english/ and http://en.wikipedia.org/wiki/Skewness [disclaimer].

Skills Transfer
See 'Capacity Building' Above.

Social Rights
Social Rights

economic

cultural rights

the first generation refers to civil and political rights;
the second generation comprises economic, social and cultural rights; and,
the third generation refers to collective rights.

Social rights are primarily aimed at ensuring individuals a specified standard of living, without discrimination. Social rights include the right to social security; the right of families, of mothers before and after childbirth, and of children to special assistance and protection; the right to an adequate standard of living, including the right to adequate food, clothing and housing; and the right to health.

Contrary to civil and political rights, which are immediately applicable and essentially based on the prohibition of States to do something, such as resort to torture, take actions that curtail freedom of speech, freedom of religion, or the right to vote, social rights tend to be considered as requiring States to take active and specific measures, such as legislation, policies or programmes, in order for those rights to be realised. Social rights are regarded as progressive: “full economic, social, and cultural rights can be achieved only gradually. Resources and time may be required” [1], though it is also clearly stated that full rights should be reached over time, and that States have the legal obligation to take immediate and continued action to do so. In addition, any action, whether legal or political taken to diminish existing protections and levels of realisation of these rights should be prohibited.

Human rights are indivisible, interdependent and interrelated, and thus the fulfillment and protection of one right affects that of others. This is true when all human rights are considered, as well as for specific categories of rights.

For instance, social rights are closely linked to economic and cultural rights in so far as the promotion of a minimum standard of living is strongly related to the right to work, the protection of property, and the right to education. Just as the distinction between civil and political rights is sometimes blurred, the difference between economic, social, and cultural rights is not always obvious. For instance, different experts regard the right to education as a social, economic or cultural right.

In international human rights law, social rights are protected by the International Covenant on Economic, Social and Cultural Rights (CESCR). The CESCR, drafted in 1966, entered into force in 1976 and is monitored by the Committee on Economic, Social and Cultural Rights, which is composed of independent experts appointed by the United Nations.

The Committee on Economic, Social and Cultural Rights is responsible for monitoring the implementation of the Covenant by its States Parties, which are required to submit regular reports on how they are implementing these rights. Such information is provided through self-reporting and thus might be limited. The reports provided are examined by the Committee, which then issues “concluding observations” in which it addresses its concerns and recommendations. To date, the Committee is not enabled to consider individual complaints against State Parties, though a draft Optional Protocol, under consideration, could provide the Committee with the jurisdiction to do so [2].

1. McChesney, A., Promoting and Defending Economic, Social and Cultural rights, AAAS/HURIDOCS, Washington DC, 2000, p. 18.

2. For further information, please refer to the website of the OHCHR, www.ohchr.org/english/bodies/cescr/index.htm.

SRTC
Statistical Research and Training Center (SRTC)

The Center's functions and responsibilities are to:

Develop a comprehensive and integrated research and training programme on theories, concepts and methodologies for the promotion of the statistical programme;
Undertake research on statistical concepts, definitions and methods;
Promote collaborative research efforts among members of the academic community, data producers and users;
Conduct non-degree training programmes to upgrade the quality of statistical manpower base in support of the needs of the statistical system; and
Provide financial and other forms of assistance to enhance statistical research and development.

Stages model
Stages Model

Lasswell

Broadly speaking, such stages include [1]:

The identification of policy problems or issues, through demands for action;
Agenda-setting, or focusing on specific problems/issues (this stage is sometimes merged with the previous one);
The formulation of policy proposals, their initiation and development, by policy-planning organisations, interest groups, and/or the executive or legislative branches of government;
The adoption of and rendering legitimate of policies through the political actions of government, interest groups, political parties;
The implementation of policies through bureaucracies, public expenditures, and the activities of executive agencies; and,
The evaluation of a policy’s implementation and impact.

By breaking the policy process up into different stages, and thus clarifying the stages that are required for a policy to be born and exist, this model has several strengths and limitations.

Its major strength is that it reduces the complexity of policy-making to manageable, analytical units, facilitating understanding. By separating the process into a series of clear and identifiable steps, one can focus on the distinct procedures and activities necessary to develop a policy, instead of losing oneself in the intricacies of the overall policy process. This model does not primarily focus on the actors and institutions involved in policy-making but rather emphasises the fact that policy-making, as a comprehensive process, cuts across and sometimes links this variety of actors and institutions (i.e. the executive and the legislative branches of government and the courts, civil society, intergovernmental bodies, etc.). As noted by Porter and Hicks in one of their papers on the process of policy formulation, by shifting attention to the process and its component steps, this model “transcends the boundaries of specific institutions and points to the ways in which individuals and groups interact across them” [2]. Entire sections of the study of policies, particularly public policies, have developed over time, based on this model.

Although this model is extremely important in the policy-making literature and is considered a traditional approach by many authors, it has also been subject to criticism. The main criticism focuses on the so-called linearity of the model.

The stages model is usually viewed as presenting the policy process in terms of a policy cycle, where the last stage of the process, “evaluation,” overlaps with the first, “problem identification”; where each step is considered as temporally and functionally distinct; and where different sets of actors are associated with different stages and periods of time. Critics say that this is extremely misleading since, although policy-making may well proceed in stages, it is not linear; the entire process is in no way automatic and can change directions or even cease to be at any point. Reality is not a fixed sequence: the “formulation” stage may be an opportunity for a new “identification of problems,” the “implementation” phase usually requires a redefinition of the “formulation” of the policy proposal, and the “evaluation” can lead to a new “problem identification.” Far from being linear, the policy process is dynamic and, at times, even chaotic. For instance, a policy can end without having been subject to evaluation, and another can be implemented before having been formally or legally adopted.

It has been said that the stages model gives the illusion that policy-makers arrive at a decision through a rational and systematic approach to problem-solving: defining the problem, analysing alternative solutions, adopting a solution, and testing and evaluating that solution. But policy-making only rarely follows this pattern. A vast number of players is usually involved in the policy process, and this tends to result in a process in which decisions are made collectively, often after resolving conflicting interests by bargaining.

1. Porter, R.W., with Hicks, I., Knowledge Utilisation and the Process of Policy Formulation: Toward a Framework for Africa, USAID, Washington DC, 1995, p. 8.

2. Op. cit., p. 9.

Standard Deviation
Variance, Standard Deviation, and Standard Error

variance

standard deviation

parameters

dispersion

In relation to data, the variance and the standard deviation are statistics that measure how widely spread the values in a dataset are. These are highly useful statistics in that they can be used to calculate a standard error, which can then be used to create confidence intervals or margins of error for sample means and other statistics.

To distinguish the standard deviation and variance that are statistics from the parameters that they estimate, we will call the standard deviation statistic the sample standard deviation, or s, and the variance statistic the sample variance, or s².

Both the sample variance and the sample standard deviation are always non-negative. If the data points are all close to the mean, then the sample variance and the sample standard deviation are close to zero. If many data points are far from the mean, then the sample variance and the sample standard deviation are far from zero. If all the data values are equal, then the sample variance and the sample standard deviation are both zero.

The formula for the sample variance, or the variance of a simple random sample of data, is as follows:

Source: Wikipedia [disclaimer]

s²

simple random sample

mean

y_i

sample standard deviation

The standard error (se) balances the dispersion associated with the underlying population and the error associated with the sampling process. It is derived by dividing the sample standard deviation, s, by the square root of the sample size, . We can think of the standard error as measuring how precisely we have estimated the population mean, or another parameter, via the sample mean, or another statistic. As the sample size gets bigger and bigger, the standard error will shrink, reflecting the fact that our estimate for the mean, or another statistic, will become more and more precise.

Calculation of the sample variance, sample standard deviation, and the standard error in the case of a complex sample design is described at http://epubs.surrey.ac.uk.

Standard Error
See 'Standard Deviation' Above.

Statement
See 'Questionnaire' Above.

Statement Taker
See 'Interviewer' Above.

Statistic
Parameters and Statistics [1]

A parameter is a numeric quantity, usually unknown, that describes a certain population characteristic. For example, the population mean is a parameter that is often used to indicate the average value of a quantity. Other, more concrete examples are:

The true “average” height of adult human males.
The number of individuals in Mexico that have paid a bribe to a police officer.
The median income in Sierra Leone.

Parameters are often estimated since their value is generally unknown, especially when the population is large enough that it is impossible or impractical to obtain measurements for all people. For example, it would be impossible to line up all adult human males on the planet and obtain their heights with perfect measurement, therefore, the true mean height of adult human males can only be estimated, not known.

Parameters are normally represented by Greek letters. The most common parameters are the population mean and variance, represented by the Greek letters μ and σ², respectively.

A statistic is a quantity, calculated from a sample of data, used to estimate a parameter. For example, the average of the data in a sample is used to give information about the overall average in the population from which that sample was drawn. Other examples include:

The “average” height of a random sample of 1,000 adult human males.
The percentage of individuals in a random sample of 10,000 adults in Mexico that have paid a bribe to a police officer, multiplied by the size of the adult population in Mexico.
The median income of a random sample of 1,000 adults in Sierra Leone.

It is possible to draw more than one sample from the same population, and each sample will have its own value for any statistic used to estimate a particular parameter. For example, the mean of the data in a sample is used to give information about the overall mean in the population from which that sample was drawn. But the sample means for two independent samples, drawn from the same population, will not necessarily be equal. Each sample mean is still an estimate of the underlying population mean. Such possible variations amoung estimates from different samples is called sampling error.

Statistics are usually represented by Latin letters with other symbols. The sample mean and variance, two of the most common statistics derived from samples, are denoted by the symbols and s², respectively.

1. This definition based on entries found at www.cas.lancs.ac.uk/glossary_v1.1/basicdef.html and http://library.thinkquest.org/10030/3smodpas.htm.

Statistics
Statistics [1]

Statistics

Statistical methods can be used to summarise or describe a collection of data; this is called descriptive statistics. In addition, patterns in the data may be modeled in a way that accounts for randomness and uncertainty in the observations, to draw inferences about the process or population being studied; this is called inferential statistics. Both descriptive and inferential statistics can be considered part of applied statistics. There is also a discipline of mathematical statistics, which is concerned with the theoretical basis of the subject.

The word statistics is also the plural of statistic (singular) which refers to the result of applying a statistical algorithm to a set of data, as in employment statistics, accident statistics, etc.

1. Derived from http://en.wikipedia.org/wiki/Statistics (accessed 31 December 2006) [disclaimer].

Statistical Research and Training Center, Philippines
See 'SRTC' Above.

Stem and Leaf plot
Stem and Leaf Plot [1]

histogram

In a stem and leaf plot, the data are arranged by place value. The digits in the largest place are referred to as the stem and the digits in the smallest place are referred to as the leaf (leaves). The leaves are always displayed to the right of the stem. An example follows:

Test Scores Out Of 100 Points (22 students)

Scores: 50, 50, 52, 58, 58, 61, 64, 64, 67, 68, 72, 74, 76, 78, 78, 79, 83, 85, 92, 92, 96, 98

Depending on the data, each stem is displayed 1, 2, or 5 times. When a stem is displayed only once (as on the plot shown above), the leaves can take on the values from 0-9.

The stem shows the “tens” and the leaf . At a glance, one can see that four students got a mark in the 90s out of 100; two students received the same mark of 92; no mark below 50 was received; no mark of 100 was received. When you count the total number of leaves, you know how many students took the test. A stem and leaf plot organises the information so that one can elicit specific information “at a glance” instead of having to sift through and analyse a long string of information.

To compare two sets of data, you can use a “back to back” stem and leaf plot. For instance, if you wanted to compare the scores of two sports teams, you would use the following stem and leaf plot:

Sports Team Scores: the Tigers and the Sharks

Tigers Sharks
0 3 7 9 |3| 2 2
2 8 |4| 3 5 5
1 3 9 7 |5| 4 6 8 8 9

The tens column is now in the middle and the ones column is to the right and left of the stem column. You can see that the sharks had more games with a higher score than the Tigers. The Sharks only had two games with a score in the 30s. The Tigers had four games, a 30, a 33, a 37 and a 39. You can also see that the Sharks had the highest score of all - a 59, compared to the Tigers with a 57.

Stem and leaf plots also enable you to find medians, determine totals, and determine the modes.

1. This definition is based on http://davidmlane.com/hyperstat/A28117.html and http://math.about.com/library/weekly/aa051002a.htm (accessed 27 December 2006).

Stratification
Stratification [1]

Stratification

sampling

random sampling

Stratification often reduces sampling error. It can produce a weighted mean that has a smaller standard error than the analogous mean of a simple random sample of the population. The error reduction is likely to occur when the phenomenon under study differs from one stratum to another, for example, if the studied variable is correlated with the age, strata such as “under 20,” “20 to 40,” “40 to 60,” “over 60” will improve the estimates accuracy.

1. Based on the Wikipedia definiton for stratified sampling (accessed 29 December 2006) [disclaimer].

Streams model
Streams model

multiple streams model

Kingdon

problem

policy

political

As noted by Porter and Hicks, this model aims to explain why some issues and problems become prominent in the policy agenda and are eventually translated into concrete policies, while others never achieve that prominence. Kingdon’s starting point is the "garbage can model" of policy-making, developed in 1972 by Cohen, March, and Olsen. This model contradicts the rational approach to decision-making, claiming that policies are not the product of rational actions, because policy actors rarely evaluate many alternatives for action and because they do not compare them systematically.

The stream of problems. o The rationale behind this stream is that a given situation has to be identified and explicitly formulated as a problem or issue for it to bear the slightest chance of being transformed into a policy. A situation that is not defined as a problem/issue, and for which alternatives are never envisaged or proposed, will never be converted into a policy. The feeling that a current or foreseen situation is wrong and that something should, and can, be done to modify and/or improve it is thus a prerequisite for turning an issue into a policy. Moreover, it is necessary to be able to demonstrate that the problems mentioned can actually be attributed to causes within human control and thus that action can be taken to change the situation.
The stream of policies. The second stream used to explain how an issue rises or falls on an agenda has to do with the stream of policies. This stream is concerned with the formulation of policy alternatives and proposals. New policies will never be shaped if there are no ideas or policy proposals on which they can be based and developed. An important aspect of the streams model developed by Kingdon is linked to the idea that such proposals and solutions, which must be technically feasible, are not initially built to resolve given problems; rather they float in search of problems to which they can be tied. A variety of actors can participate in the elaboration of such solutions and alternatives, and in the drafting of proposals for policy reform.
The stream of politics. Although they take place independently of the other two streams, political events, such as an impending election or a change in government, can lead a given topic and policy to be included or excluded from the agenda. Indeed, the dynamic and special needs created by a political event may change the agenda. In the political stream, consensus is usually obtained as a result of bargaining rather than persuasion. Thus, more attention is paid to assessing the costs and benefits of a policy proposal than to underlining its analytical importance and relevance.

As mentioned above, these three streams are separate and independent; problem recognition, the formulation of policy proposals, and political events each has its own dynamic and pace. As such, no stream is decisive to the overall policy process, though all streams are important. It is when they meet and coincide that an issue is transformed from a mere topic and/or problem into a concrete policy, that is, a compelling problem is linked to a plausible solution that meets the test of political feasibility [2]. For example, supporters of a given policy reform take advantage of a political context that favours and seeks new ideas and approaches, claiming that their proposal for reform is also a solution to a previous problem. In this instance, there is a complete linkage between the three streams, which increases the chances for an issue to become a policy.

However, it is not always necessary for all three streams to meet simultaneously for a policy to develop. In some cases, partial couplings, the convergence of two of the streams, are sufficient, though the whole policy-making process is more uncertain. Kingdon argues that policy entrepreneurs play a key role in connecting the streams, and that there are different types of couplings. Couplings can be more or less ‘tight’ or “loose,” depending on the degree to which streams, though independent, depend on each other for an issue to develop into a concrete policy [3].

Contrary to the stages model, the streams model does not picture the policy-making process as one that involves steps and stages. Rather, it views the policy process as the result of the intersection of at least two independent streams at one time. In this model, there is no chronological sequence or priority among the streams. Streams act and react according to their own logic, until a window of opportunity is opened and two or more streams coincide and become a policy.

The major strength of this model is that it recognises that the policy process is fluid and non-linear, and that it involves a vast number of actors and forces. It also explains how a given issue becomes a specific policy—or not.

1. For more information, see Boussaguet, L., Jacquot, S., et Ravinet, P., Dictionnaire des politiques publiques, Presses de la Fondation Nationale des Sciences Politiques, 2004, p.217-225.

2. Op. cit., p.19.

3. For further information, see Lemieux, V., L’étude des politiques publiques : les acteurs et leur pouvoir, 2e édition, Les Presses de l’Université Laval, Canada, 2002.

Survey
Survey [1]

survey

survey instrument