Comparing Approaches to Measuring Parental Nonstandard Work Schedules

Background and Purpose: The development of 24/7 economies has led to a growing body of research on the prevalence and consequence of parental nonstandard work schedules (NWSs; work outside a traditional 9-to-5, Monday to Friday time span). Some surveys utilized in this research domain ask respondents to select only one type of schedule that best describes their primary work schedule (e.g., standard vs nonstandard), while others allow respondents to indicate whether they work both types of schedules. A thought-provoking study by Dunifon and colleagues (2013) provided compelling evidence that allowing respondents to choose more than one schedule doubles the prevalence of maternal NWSs, compared to requiring respondents to report only one schedule. Building on the previous study, this study examined whether the different measurement approaches influence our understanding about the consequence of parental NWSs, by investigating the association between parental work schedules and health outcomes.

Methods: Using a multi-stage sampling method, we gathered survey data from a sample of parents with young children (aged 5-6 years) in Hong Kong (N = 732). When measuring work schedules, respondents were allowed to choose multiple options from the following categories: 1) standard (8 am and 6 pm), evening (6pm-midnight), night (midnight-8am), rotating shifts, and irregular hours. They were then asked to select only one schedule that best describes their primary work schedule. We first estimated the prevalence of parental NWSs to examine whether prevalence estimates differ by measurement approaches. We also conducted linear regression to explore whether regression estimates regarding the association between NWSs and health outcomes (measured by self-rated health items) differ across measurement approaches.

Results: When respondents were instructed to choose only one schedule, 20.5% of the sample reported that they worked a NWS. When they were allowed to indicate multiple schedules, the prevalence of NWSs was estimated to be 36.3%, as 23.6% of the total sample (equivalent to 65.0% of parents working NWSs) indicated that they regularly worked both standard and nonstandard hours. Regression results also revealed differences based on measurement approaches. When the work schedule variable was constructed based on items requiring respondents to select only one schedule, the association between working NWSs and health outcomes was non-significant. In contrast, when respondents were allowed to choose multiple schedule options, working NWSs had a significant association with adverse health outcomes. Additional analyses uncovered that this discrepancy was mainly due to findings that parents working both standard and nonstandard hours were particularly vulnerable to adverse health outcomes, while the health outcomes of those working nonstandard hours only were not significantly different from those working standard hours only.

Conclusions and Implications: This study offers evidence that the way we measure work schedules substantially influences our understanding about the consequences and the prevalence of NWSs. Given that many parents have jobs involving both standard and nonstandard hours, mutually exclusive categorization of standard vs. nonstandard may not capture the complex landscape of NWSs. Future studies using diverse measurement approaches are warranted.