In our latest guest blog, Richard Laux, Deputy Director, Data and Analysis, at the Race Disparity Unit, discusses the proposed changes to the Ethnicity facts and figures website, for which a consultation was launched today. The consultation will gather views about how users of Ethnicity facts and figures understand the drivers and factors behind disparities, minimising the risk of misinterpretation and incorrect conclusions being drawn – and in turn adding value, as defined in the Code of Practice for Statistics.
The consultation is in response to Action 8 in the recently published Inclusive Britain report.
The Ethnicity facts and figures website was launched in October 2017. The intention was to increase awareness of disparities in outcomes and experiences between ethnic groups to stimulate debate and, in turn, action to reduce disparities. User research pointed to the importance of presenting the ‘raw’ unadjusted data and descriptive commentary, with no modelling and no attempt to ‘explain’ disparities; also, to adopt a common format and layout throughout.
Over 4 years later the context in which people think about ethnic disparities, and hence the user need for information, has developed. The research conducted to inform the report of the Commission on Race and Ethnic Disparities (to which Inclusive Britain was the Government’s response) pointed to a need to try to better focus on understanding disparities and outcomes for specific ethnic groups.
We are proposing three changes to the website to achieve this new goal. The consultation itself, and all three proposed changes, are firmly rooted in two of the Code’s principles that are about value:
- relevance to users -”users of statistics and data should be at the centre of statistical production; their needs should be understood, their views sought and acted on, and their use of statistics supported” (V1).
- innovation and improvement – “statistics producers should be creative and motivated to improve statistics and data, recognising the potential to harness technological advances for the development of all parts of the production and dissemination process”. In particular, “users should be involved in the ongoing development of statistics and data, exploring and testing statistical innovations, so that the statistics remain relevant and useful” (V4.3).
The first proposed change is to provide different levels of information for different measure pages. For example, we think it is sensible to reduce the amount of commentary, charts and additional analytical splits for pages that are not accessed frequently. And we think we can make some data more accessible by combining measure pages on related topics, such as pupil performance at each Key Stage. We see this as being consistent with the following aspects of the Code:
- Statistics, data and related guidance should be easily accessible to users. The needs of different types of users and potential users should be considered when determining ways of presenting and releasing the statistics and data (V2.2)
- Statistics, data and explanatory material should be relevant and presented in a clear, unambiguous way that supports and promotes use by all types of users (V3.1)
- Statistics should be accompanied by a clear description of the main statistical messages that explains the relevance and meaning of the statistics in a way that is not materially misleading. They should be illustrated by suitable data visualisations, including charts, maps and tables, where this helps aid appropriate interpretation of the statistics (V3.2)
The second proposed change relates to those measure pages which present data according to a binary classification – for example, ‘white’ and ‘other than white’. Such data splits are sometimes shown because small sample sizes or populations do not allow for data to be shown for more than 2 ethnic groups, but this type of data is not useful because it does not distinguish between diverse experiences within or across particular ethnic groups. We are proposing to work with the relevant Departments to provide more detailed ethnicity classifications before updating these measure pages. We see this as being consistent with the following aspects of the Code:
- Statistics, data and metadata … should be released at the greatest level of detail that is practicable to meet user needs (V2.4)
The third proposed change is to provide additional analysis and context about priority topics through short summaries and links to wider government and academic research. For example, we may provide: the results of regression analysis (where appropriate) to demonstrate the extent to which disparities may be accounted for by factors other than ethnicity, research into geographic variations, contextual information, such as ‘pipeline’ effects, and links to other research, including academic papers and qualitative investigations.
This will help users to understand the drivers and factors behind disparities – minimising the risk of misinterpretation and incorrect conclusions being drawn, and providing better evidence for targeting interventions and resources. We see this as being consistent with V3.2 (as above) and also:
- Comparisons that support the appropriate interpretation of the statistics, including within the UK and internationally, should be provided where useful. Users should be signposted to other related statistics and data sources and the extent of consistency and comparability with these sources should be explained to users (V3.3)
We are also working on a further refinement of our statistical activity underpinning the website. This does not impact on users, so is not part of our consultation. But we hope that the use of new technologies – Reproducible Analytical Pipelines – will simplify the way we update data on the website, freeing up our resources to conduct value added work such as analysis to help understand disparities, and reducing the burden on the Departments that provide us with the data. We think this is consistent with the following aspects of the Code:
- Statistics producers should keep up to date with developments that can improve statistics and data (V4.1)
- Statistics producers should be transparent in their approach to monitoring and reducing the burden on those providing their information, and on those involved in collecting, recording and supplying data (V5.5)
If you have views on the presentation of ethnicity data, I encourage you to complete our consultation.