There is a large body of research validating self-report medical data with that obtained through linkage, for example (1–3). There is however far less research that has investigated the validity of educational data self-reports, with studies restricted to small samples (4) or single courses (5). The accuracy of official measures of education such as the UK national Pupil Database is cited as one of the main advantages to conducting data linkage in cohort studies, and therefore determining the accuracy of self-reports when compared to these official measures is of critical importance for assessing the benefits of data linkage. Furthermore, the direction and magnitude of mis-reporting has been found to vary with individual characteristics(5) which will means that measurement error will contain not just random noise but also directional bias.

ALSPAC holds data on G1 participants’ education via data linkage only. Retrospective collection of participants education through self-report responses to questionnaires would allow the accuracy of this data to be compared to the linked data. Given the rich data that ALSPAC holds on its participants it may also be possible to investigate how misreporting varies across a range of individual characteristics including sex, intelligence, and family socioeconomic position. We note however that this analysis into misreporting may be restricted due to the non-random nature of cohort attrition in ALSPAC. Below we highlight some of the research questions that could be addressed through the collection of self-report education data collected by ALSPAC.

The importance of this data collection is demonstrated by recent changes in data linkage surrounding data sharing. The NPD recently decided to temporarily withdraw all data linkage due to data sharing and confidentiality issues, highlighting the risk that because linked data is not owned by cohort studies it can be pulled at short notice. Collecting self-report education data will therefore also provide ALSPAC with a safety net in the form of its own accurate educational records in the worst-case scenario that NPD data linkage is revoked permanently. While this self-report data will be only a sub-sample of the cohort due to attrition, it will ensure a larger sample than collection at a future occasion due to continued attrition.

As part of this data collection we will clean, code and deposit all data with full accompanying documentation and code for use by others.

1. Hafferty, J. D. et al. Self-reported medication use validated through record linkage to national prescribing data. J. Clin. Epidemiol. (2018). doi:10.1016/j.jclinepi.2017.10.013
2. Comino, E. J. et al. Validating self-report of diabetes use by participants in the 45 and up study: A record linkage study. BMC Health Serv. Res. (2013). doi:10.1186/1472-6963-13-481
3. Mars, B. et al. Using Data Linkage to Investigate Inconsistent Reporting of Self-Harm and Questionnaire Non-Response. Arch. Suicide Res. (2016). doi:10.1080/13811118.2015.1033121
4. Sticca, F. et al. Examining the accuracy of students’ self-reported academic grades from a correlational and a discrepancy perspective: Evidence from a longitudinal study. PLoS One (2017). doi:10.1371/journal.pone.0187367
5. Rosen, J. A., Porter, S. R. & Rogers, J. Understanding Student Self-Reports of Academic Performance and Course-Taking Behavior. AERA Open 3, 2332858417711427 (2017).