Studying Differences of Household Weekday and Weekend Activities

A desired activity-based travel demand modeling framework should be able to address both weekday and weekend activities. However, a literature review shows that previous research efforts have mostly focused on weekday, not weekend, activities, and that little or no research exists to quantify the differences between the two. The best knowledge to date is limited to weekday and weekend activities that start at different times of the day and have different participation rates. This paper aims to fill the gap by studying the differences between weekday and weekend activities in Calgary, Canada, in terms of participation rates, starting times, duration, and inferred location choices. First, statistics related to these attributes were computed for 10 types of weekday and weekend activities (these were found to differ). Second, log-rank and Wilcoxon tests were used to prove further that common types of weekday and weekend activities tend to follow different survival functions. Third, best-fit duration models were explored for each type of weekday and weekend activity and compared with each other. It was found that Weibull and log-normal were chosen as the best-fit models for nearly all weekday and weekend activities. The best-fit duration models for the same types of weekday and weekend activities (e.g., shopping) were found to be different in either underlying distribution or estimated parameters. This study clearly shows that the weekend activities differ from their weekday counterparts and suggests that they be treated separately in activity-based modeling frameworks.