Thursday 3rd December 2020, 13:10–13:25 (Australia/Sydney), Zoom Breakout Room 1
When administering large scale surveys, it can be tempting to include as many questions or instruments as possible to allow complex analyses involving many variables. However, there is inevitably a trade-off between survey length and responder engagement. On the other hand, shorter instruments are generally less reliable. Ideally, measures are as short as possible while still remaining psychometrically valid. In this session we will present the findings from a recent study involving the Longitudinal Surveys of Australian Youth (LSAY), in which we introduced four new measures of ‘soft skills’ in a survey of nearly 5,000 Australians aged 15-19. We will cover all stages of the instrument development process, beginning with a theoretical framework which guided the identification of candidate items, to survey administration and validation through principal components analysis and generalized linear mixed modelling with longitudinal outcomes.
Dr Forrest has worked at the National Centre for Vocational Education Research (NCVER) since 2016. He is a Data Analyst in the National Surveys branch, where he primarily works on the Longitudinal Surveys of Australian Youth (LSAY).