tailieunhanh - Child mental health and educational attainment: multiple observers and the measurement error problem

The PhD in Biostatistics provides training in the theory of probability and statistics in biostatistical methodology. The program is unique in its emphasis on the foundations of statistical reasoning and in requiring its graduates to complete rigorous training in real analysis-based probability and statistics, equivalent to what is provided in most departments of mathematical statistics. PhD candidates are required to pass a comprehensive written examination covering coursework completed at the end of their first year. Research leading to a thesis may involve development of new theory and methodology, or it may be concerned with applications of statistics and probability to problems in public health, medicine or biology. Applicants should have completed undergraduate work in the biological, physical or social sciences. | Child mental health and educational attainment multiple observers and the measurement error problem David Johnston Monash University Carol Propper Imperial College and University of Bristol Stephen Pudney Institute for Social and Economic Research University of Essex Michael Shields Monash University and University of Melbourne No. 2011-20 August 2011 INSTITUTE FOR SOCIAL ECONOMIC RESEARCH E-S-R-C economic SOCÍẬL RẼS_ẸẠỆỘH COUNCIL Non-technical summary Child mental health is an important social and economic issue not only because of its implications for the wellbeing of children but also because mental health problems have been linked with poor educational achievement and consequent lifetime disadvantage. However research on child mental health is problematic in part because of the difficulty of observing and measuring a child s state of development and mental health. Research is typically based on either diagnostic data from a clinical setting or on large-scale surveys which ask parents or teachers to assess the child s mental health using a structured questionnaire. The former approach is generally based on small unrepresentative groups and is hard to generalise to the wider population of young people. The survey approach suffers from the problem of measurement error -parents and teachers may not be accurate observers and reporters of the child s behaviour and mental state. More seriously these non-expert observers may be not only inaccurate but systematically so either because they have only a partial picture of the child s behaviour or because they are subject to bias in some way. We evaluate these problems by using unusually rich survey data which provide assessments from parents teachers and children themselves together with an overall expert assessment which approximates the clinical diagnostic process. We find evidence that parents teachers and children are all biased reporters of children s mental health but that using expert quasi-diagnoses as a yardstick