Glossary >> Glossary of statistical terms Error (see also bias)  The difference between the observed or estimated value of an indicator and its “true” value. Errors may be random or systematic. Systematic errors are called “biases”. Random or sampling error can be thought of as “the difference between a sample and the population from which the sample is derived” and balances out on average. Sample surveys are nearly always affected by sampling error. As the sample size increases, the sampling error decreases. The total error between the estimated value of an indicator and its true value is a combination of sampling error and bias. The fact that error is common in working with statistics does not mean that statistical indicators are not useful. With resources and appropriate methodologies near “perfect” estimates of true population values can be estimated. When comparing indicators across territories or over time, we must interpret differences in observed values with caution—they could be the result of such error. Gini coefficient A common measure of the distribution of a country’s wealth, income or private/household consumption, ranging from 0 to 1 or 0 to 100, where 1 or 100 denotes complete inequality and 0 denotes complete equality. Indicator  Information that indicates a state or level of an object, event or activity. It provides an indication of prevailing circumstances at a given place and a given point in time. Often based on some form of quantification (e.g., proportion of children immunized) or qualitative categorization (e.g., a treaty ratified/not ratified). In the context of this work, an indicator can be considered as a human rights indicator if it can be related to human rights norms and standards, addresses and reflects human rights principles and concerns, and is used to assess and monitor the promotion and implementation of human rights. Indicator reliability (see also indicator validity) The consistency in the value of a variable/indicator reported by different data producers when using the same method and data source. Indicator validity The soundness of a variable/indicator in measuring what it seeks to measure. If someone who weighs 200 pounds steps on a scale 10 times and gets readings of 15, 250, 95, 140, etc., the scale is not reliable. If the scale consistently reads “150”, then it is reliable, but not valid. If it reads “200” each time, then the measurement is both reliable and valid. Metadata  Data that describe the characteristic details of an indicator. They usually include information on the definition, rationale, method of computation, data collection and source, disaggregation, periodicity, comments and limitations of that indicator. Performance indicators  In the context of results-based management (RBM) of development intervention, a country programme or any other project carried out by an organization, performance indicators refer to quantitative or qualitative variables that allow the verification of changes resulting from the intervention or show results relative to what was intended or planned. Proportion  A ratio where the denominator is a quantity that represents the given population group and the numerator is only a subset of that population group. For example, the proportion of farmers availing of extension services is calculated as the number of farmers availing of extension services divided by the total number of farmers. Furthermore, if the proportion is multiplied by 100, it becomes a percentage. 172 Human Rights Indicators