Dear Siobhan
Women in Northern Ireland Statistics
I am writing to you following our review of the Northern Ireland Statistics and Research Agency’s (NISRA) Women in Northern Ireland Statistics, against the Code of Practice for Statistics. I am pleased to confirm that these statistics should continue to be designated as National Statistics.
The publication is one of several thematic bulletins concerning the labour market in Northern Ireland and covers the different experiences of women and men in the labour market. Since the publication of the New Decade, New Approach Deal in 2020, which acts as a basis for restoring the Northern Ireland Executive, the Department for Communities has been leading on the development of several Executive Social Inclusion Strategies for Northern Ireland. This includes the Gender Equality Strategy, which has led to heightened interest in the Women in Northern Ireland statistics.
The production of these statistics demonstrates a level of statistical curiosity in the statistics team, which was reflected in our conversation with the team. The team told us it has been working with the relevant policy teams to develop indicators to support the emerging Gender Equality Strategy, as well as other policies resulting from New Decade, New Approach. As well as meeting these data needs, the team carries out regular reviews of user queries to remain alert to areas of emerging interest and to ensure the statistics remain relevant to users. For example, the team plans to expand future publications to cover measures of work quality such as security and flexibility of work.
This engagement is further supported by the team’s latest annual meeting with the labour market statistics user group, in which the team presented its plans for the future of the Women in Northern Ireland statistics. Through this engagement, the statistics team has sought feedback on the name of the bulletin, given it covers men and women in the labour market, and other thematic topics which would be useful to explore in more detail. We are pleased to see the statistics team engage with users to make these sorts of decisions and we would encourage the team to continue to work with users to explore the demand for further thematic analyses.
The quality information in the bulletin is written in plain English, which can be understood by a wide audience. A particularly good example of this is the explanation of how it is possible for the number of unemployed to increase and the unemployment rate to fall during the same period. This aids users’ understanding and interpretation of the statistics. We are also pleased to see the clear summary of how population weights are used to produce estimates for Northern Ireland and an explanation of the thresholds used to determine whether the relevant breakdowns from the Labour Force Survey are suitably robust for inclusion in the bulletin.
The bulletin acts as a good example of how to present uncertainty in statistical releases. There is an explanation of what confidence intervals are and how they have been used. The bulletin also prominently states that the statistics are subject to sampling error and links to definitions as well as a specific document on uncertainty in the Labour Force Survey. To further build on the presentation of uncertainty, the statistics team should consider publishing the confidence intervals for all of the breakdowns where possible, rather than only a selection of them.
The statistics team has previously used shaded cells in the data tables to indicate smaller sample sizes and highlights that these result in imprecision. It told us that it is reviewing the use of shading in the tables as part of a wider review of accessibility of outputs. We are pleased to see the efforts taken to present uncertainty and welcome the team’s work to balance these with accessibility concerns. We are also pleased to hear that NISRA is exploring the role of Reproducible Analytical Pipelines in its work to develop a framework for accessible tables and bulletins. This is a great example of using innovative ways to address development goals.
We would like to thank the team for its cooperation throughout the review process. I am copying this letter to Cathryn Blair, acting Head of Economic and Labour Market Statistics for NISRA and Brian Green, Director of Sources in NISRA.
Yours sincerely
Mark Pont
Assessment Programme Lead