New Methods in Time-Use Data Collection

• Activity Number: 434 - Contributed Poster Presentations: Business and Economic Statistics Section
• Type: Contributed
• Date/Time: Wednesday, August 10, 2022 : 10:30 AM to 12:20 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Sophia Arabadjis* and Stuart Sweeney
• Companies: University of California Santa Barbara and University of California Santa Barbara
• Keywords: Type II Error; Microsimulation; Time Use; Power; Sampling

Abstract

Time use studies have a long and storied history spanning centuries and disciplines. In modern times, time use studies are collected by many official government statistics agencies to gain insights into differences in sub patterns of work and leisure across relevant social strata. However, collection of time-use data in low-income countries and sub-literate populations using traditional methods remains a challenge. We have pioneered a new method of time-use data collection using in-situ audio recorded activity clips which we have deployed to study time poverty in rural Senegalese agriculturist households. Our new method complicates traditional statistical methods in time use, as the intra-household correlation violates independence assumptions and interpretation of subgroup differences as well as the underlying mechanism that generates the activity structure. In this study, we use simulated activity times with known differences across sub populations based on empirical measures from a case study to assess sample sizes necessary to detect differences using fixed and random time-clip sampling. We repeat this analysis using activity generated from uncorrelated activities as well as correlated activities.