III. >> M  ethodological Approaches to Human Rights Indicators >> Sources and data-generating mechanisms In conformity with human rights and international statistical standards; and Amenable to disaggregation by prohibited grounds of discrimination and by vulnerable or marginalized population group at country level. An indicator has to be simple (to understand and apply), and easily available in a timely manner to be a meaningful tool in undertaking human rights analysis and assessment. Otherwise, the opportunity cost of collecting and compiling relevant information on the indicator could become a deterrent. These factors should be taken into account in considering whether an indicator should be used in periodic reports to the treaty bodies or in the universal periodic review, or for follow-up to treaty body recommendations. The reliability of an indicator refers to its consistency in the estimate or the value of an indicator if the data-generating mechanism employed for devising it is repeated. For instance, if a question is asked for a second time to the same person and it produces an identical response, everything else being equal, then the question/response could be considered as a reliable indicator. Often, this is not the case if the question is formulated in an ambiguous manner. Moreover, the reliability of an indicator is affected by biases in data-generating mechanisms, which, inter alia, could be the result of misspecification of questions or definitions, apprehensions of the respondents, or non-representativeness of the sample.6 For an indicator to be accepted and applied as a tool in human rights analysis, it has to be based on transparent and verifiable methodology. Indicators based on haphazard information and subjective approaches to data generation are less likely to be effective or credible. Indicators are more credible when they are reliable, relevant and based on ethical and scientific principles of data collection, processing, storage and presentation (see box 12). B. Sources and data-generating mechanisms Based on a survey7 and assessment of some major attempts at and approaches to developing quantitative human rights and related indicators, one can identify at least four broad categories of data-generating mechanisms that could potentially be useful in developing indicators for use in human rights assessments. These are highlighted with representative examples and analysed for the elements that each category of data types could bring to the human rights assessment process and methodology. There are two considerations that stand out in this context. First, the sources and identified data-generating mechanisms should be suitable for assessing the compliance of State parties with international human rights treaties. As a result, the focus should be on indicators that are fact-based or use objective methods of data collection and presentation. Second, there is a need to combine different sources and data-generating mechanisms to encourage a more comprehensive and credible assessment of any human rights situation. 6. See “Bias” in the glossary of the Guide as well as in Asher, Banks and Scheuren, eds., Statistical Methods. 7. See Malhotra and Fasel, “Quantitative human rights indicators”. This survey is by no means exhaustive. It draws from some attempts at mapping and surveys of human rights and related indicators and some earlier studies, in particular M. Cain, R. Claude and Th. Jabine, “A guide to human rights data sources”, in Human Rights and Statistics: Getting the Record Straight; T. Landman and J. Häusermann, “Map-making and analysis of the main international initiatives on developing indicators on democracy and good governance” (2003); UNDP, Governance Indicators: A Users’ Guide, 2nd ed. (2007); and C. Naval, S. Walter and R. Suarez de Miguel, “Measuring human rights and democratic governance: experiences and lessons from Metagora”, OECD Journal on Development, vol. 9, No. 2 (2008). Human Rights Indicators 51