| Day-to-day variations in individuals’ activity/travel behavior have always been an attractive issue in transportation studies. While a number of studies have been focusing on the differences between weekends and regular weekdays, it seems as if the role of shoulder days have been underestimated, if not completely neglected. In this content, this study seeks to examine the daily fluctuations in activity patterns and to highlight the need for modeling the travel patterns on all days of the week. This study provides an explanatory examination of within-weekday irregularities observed in individuals’ non-mandatory activity behavior with an emphasis to highlight contrasts between shoulder days(Mondays and Fridays), and mid-weekdays(Tuesday through Thursday). Specifically, social activities are investigated. General descriptive statistics were investigated which help the analyst shed light on the existing contrasts primarily observed between the week day categories. Appropriate tests were carried out to investigate the statistical significance of the values. In the second step, which consists of advanced econometric analysis and modeling, a sample selection model was fitted to the dataset. Binary Probit and linear regression models were developed simultaneously to predict two major activity dimensions, namely participation and duration. In particular, interactive effects of explanatory variables with weekday categories were taken into consideration. The model results reveal significant interaction effects with the weekday variable in some cases, which confirms the hypothesis that activity participation varies between shoulder days and mid-week days, and this change is influenced by the socioeconomic and demographic segments. Conclusions in this study benefit relevant public sectors when making transportation policies and plans. The data applied in this study was obtained from the 2009 National Household Travel Survey(NHTS) and 2009 American Time Use Survey(ATUS)... |
PDF Full Text Request