Given that we are as yet unable to apply many of the criteria to our data on displacement associated with disasters, we have only assessed our confidence in the figures associated with conflict and violence. In doing so, we applied a common set of criteria based on: || The methodologies used to collect it || Whether it could be independently validated || The degree to which it is geographically comprehensive in terms of the extent of the conflict and associated displacement || Whether it is disaggregated by sex and age || The frequency with which it was collected || How extensively it covers the components of our data model lation and gaps to be identified, while its absence can lead to possible double-counting. TFPM in Yemen uses a similar rationale in its confidence rating to justify discarding data when location information is incomplete. We have not attempted to weight or rank these factors, nor have we assigned quantitative point values for them or generated an overall score for each source and estimate. In order to do so rigorously, we will first need to empirically test the relative significance of each of the factors. Temporal dimensions: The frequency of updates is a relative criteria. Unfolding crises and rapidly changing situations such as those in Syria, Iraq and Yemen require more frequent updates than stable and often protracted situations such as in Armenia and Cyprus. Yearly updates may suffice for some situations, but for others, it can exclude some of the shorter-term displacements. Some of the data gaps reported can be attributed to the way governments and organisations collect and disseminate data, but this is not always the case. We try to be as comprehensive as possible in our own data collection, but we may overlook some sources that may address the gaps we report. As such, our assessment reflects the level of detail of the data we were able to collect and process from various sources – not the level of detail of all the data that exists or was published by each provider. Our confidence assessment for the largest stock and new displacement figures associated with conflict is shown below in table A.10. Our assessment for the full list of countries is available on our website. Notes on IDMC’s confidence assessment criteria Data disaggregated by sex and age (SADD): The availability of SADD does not directly factor into the calculation of the number of IDPs, but it can be considered a proxy for detailed data collection practices. Geographically disaggregated data: Such data is not, per se, an absolute requirement for accurate national estimates of displacement. In many countries, however, some of the entities that collect data only have access to some regions. Geographical disaggregation allows for triangu- 110 GRID 2017 Multiple data sources: The availability of data from a number of independent sources does not guarantee higher quality or more accurate overall results. It can, however, prompt discussion of the various estimates available and the methodologies used to derive them. It also sometimes permits triangulation, which is useful in situations for which displacement estimates are highly sensitive or more susceptible to data collectors’ biases. Next steps Our confidence assessment is a work in progress, and we welcome input from partners interested in contributing to its development. For this report we assessed our confidence in all the conflict figures reported. This represents a significant increase with respect to the GRID 2016, where only 11 countries were considered. We plan to apply our criteria to all of the data we receive and analyse so that our estimates are as accurate as possible. In doing so, our data users will be made aware of the magnitude of uncertainty the data contains, and the underlying reasons for it. Notes 1. 2. 3. 4. UN, World Population Prospects, 2015 revision, available at goo.gl/TFtEjY World Bank, Country and Lending Groups, available at goo.gl/vvClzK IDMC, Global Report on Internal Displacement 2016, May 2016, p.79, available at goo.gl/VOc8OZ Benini A, Shikh Aiyob M, Chataigner P et al, Confidence in needs assessment data: the use of confidence ratings in the Syria multi-sectoral needs assessment (MSNA), a note for ACAPS and MapAction, April 2015, available at goo.gl/Vo7W01