Logo image
Back
Analysis of LSAC Time Use data: Opportunities and Challenges
Abstract

Analysis of LSAC Time Use data: Opportunities and Challenges

J Brown and Peter K Dunn
Proceedings of ACSPRI Social Science Methodology Conference
ACSPRI Social Science Methodology Conference, 2006 (Sydney, Australia)
2006
url
http://www.acspri.org.au/View
Webpage

Abstract

Sociology Statistics time use children
The longitudinal study of Australian Children (LSAC) has included time diaries to collect information on how Australian children spend their time. Consistent with the main study, diaries were collected from two cohorts in Wave 1, these being an infant cohort and a 4-5 year old cohort. The time diaries are 'light' diaries and contained information on many pre-coded activity and context domains. These include activities such as sleep, eating, travel, watching television and play as well as information about where the child was and who they were with and whether or not someone had been paid for the activity. Parents or guardians were asked to complete two diaries for their child, one on a weekday and one on a weekend. Like all time use data, the information collected from LSAC is very rich in information not only because the diaries collected data on a broad range of activities but also because of the fabric of life itself. Children and adults alike most commonly eat more than one meal a day but may only go for a walk once a week. Most commonly time use research focuses on the total durations spent in activities of interest. The often right skewed zero limited distributions that result because adults and children engage in some activities relatively infrequently are seen as problematic because linear regression cannot be applied without ignoring the assumption of normality. However, the Poisson-gamma distribution is a little known distribution that takes into account the episodic nature of the time use data. The Poisson-gamma distribution is a Poisson sum of gamma distributions. That is, there is a Poisson sum of activity episodes, each following a gamma distribution. Further, time use data opens many possibilities for the social scientist because the timing and/or the sequences of the activities may give greater insight into the nature of a particular problem rather than limiting line of enquiry to the total amount of time spent in an activity of interest. The aim of this paper is to explore the unique opportunities and challenges associated with the analysis of time use data. Examples shall be drawn from the LSAC time use data.

Details

Metrics

4 File views/ downloads
658 Record Views
Logo image