(In)Consistency of responses to self-assessed health measures and implications for biosocial research

The interplay between socio-economic circumstances and health is of major importance for individuals' lives and the human capital investment of the society. However, much of the existing research is conducted using subjective, self-reported proxies of health, which may impose significant biases. Recent advances in the large social-science surveys involve the collection of physical measurements and markers derived from biological samples, in addition to the conventional self-reported health assessments. In this project, we focus on the most popular self-reported health measures, self-assessed health (SAH). We exploit a distinctive feature of the ESRC funded national panel, Understanding Society: the UK Household Longitudinal Study (UKHLS), with longitudinal information on SAH collected twice for each respondent within waves, with different survey modes (open interview versus self-completion) taken within a short space of time. About 25% of respondents changed their responses within the same wave, with about 25% of those with inconsistent reports at UKHLS wave 2 also reporting SAH inconsistently at UKHLS wave 5. Capitalising on this feature of the survey's design along with the availability of objectively measured biomarker data and a rich set of longitudinal socio-economic data for all adult household members we are aiming to make a number of distinct contributions to social science research on health, survey measurement of health and health policy.

The proposed project focuses on reporting inconsistencies in SAH and the potential implications for the social science research and policymaking. Although SAH is one of the most popular health measures used in economics, sociology, social epidemiology, medical research and as instrument in clinical trials, research on assessing the consistency of responses to SAH questions is sparse compared to that using SAH as a measure of health. We anticipate that our project will illustrate the importance of capacity building and further development for biosocial data sources. Specifically, in this project we will address research questions on the consistency of the SAH responses, characterising the demographic and socio-economic profile of those who are more likely to provide inconsistent SAH reports as well as the role of the micro-social environment at the fieldwork (interview) stage in affecting individual responses to SAH questions. We will also address the implications of reporting inconsistency in SAH for existing research on socio-economic inequalities in health as well as on the ramifications of ill health for people's employment outcomes (such as unemployment, wages and retirement decisions). A prerequisite for scientifically-sound, evidence-based social and health policy is the availability of robust and unbiased research evidence.

Dissemination and communication with the research community will operate through conference presentations, symposia and four publications in peer-reviewed journals to draw attention to our research findings. We will achieve a good representation of our findings in discipline-specific and multidisciplinary journals. We will also work with policymakers and third-party organisations to disseminate and enhance engagement to the policy implications of our analysis and the value of robust health measurement using more objectively measured health proxies.