UK-Australia Network Promoting Longitudinal Survey Methods and Analysis

点击翻译按钮获取中文摘要

This proposal would create an international partnership network between the Institute for Social and Economic Research (ISER) at the University of Essex and the Universities of Melbourne and Queensland in Australia. The objective of this network is to build and exchange knowledge focused on the link between panel survey methods and substantive analysis in key policy areas. The funding in this proposal would cover travel, subsistence and costs associated with the organisation of a meeting, two workshops and international researcher exchange visits. The funding would not be used for research or for staff costs. Those who work trying to improve the quality of survey data generally use a framework of "Total Survey Error", which separates the different types of errors that surveys can have. These errors affect all surveys but are particularly significant on longitudinal surveys where the same people are being interviewed over time. Research into these errors generally focus on summary measures such as the means and percentages generated by the survey. However, most substantive research into important policy areas relies on types of multivariate analysis which use information about the distributions of responses within the data (regression, logit or probit models). The impact of survey error on estimates derived from such multivariate methods cannot be understood by focusing only on marginal summary measures. The results of substantive researchers using multivariate analyses may be seriously distorted by errors that arise due to the design of the survey, for example, if the way that the survey was organised influences the answers that people give in different ways. Little is known about how the design of panel surveys affects the stability of substantive answers across time, particularly in the key policy areas of employment, health, well-being, poverty and family dynamics. To our knowledge research on the causes of longitudinal inconsistencies and, most importantly, their substantive effects is lacking.This network will provide opportunities to forge long-term relationships with Australian colleagues, to exchange ideas and to promote research on these topics. The network will, first, target methods for improving data quality, both during data collection and after collection. The network will then aim to develop sustainable research collaborations focused on the effect of improved data quality in substantive research, such as measures of employment change, changes in health or well-being, or family dynamics over time. The funding from this scheme would establish the infrastructure of an international partnership which would enable researchers to travel between the UK and Australia to collaborate on research projects. These research projects will be funded separately. The funding from this scheme would establish the infrastructure of an international partnership which would enable researchers to travel between the UK and Australia to collaborate on research projects. These research projects will be funded separately. The level of commitment to the project shown by the international partners is demonstrated by the level of proposed co-funding. The international partners are pledging direct financial contributions of up to £21,000 over the three year project.