Work-family conflict (WFC) is an important line of inquiry in organizational behavior and human resource management research. The topic is relevant to the computing and communication field not only because modern communication technologies allow for more integration of work and family roles than ever before15 but because recent advances in computing technology offer new ways to respond to and understand WFC. WFC has been empirically associated with employees' job and life dissatisfaction, poor physical and psychological health, and rising voluntary turnover rates and work stress.5 WFC has also been found to have a negative correlation with various aspects of organizations, including performance, commitment, psychological contract, and even strategy planning.20 Our analysis of American Community Survey and Census data21 from 2015 found over 75.2% of males and over 60.5% of females in married couples have their own earnings, with dual-income families emerging as the predominant family structure in the U.S.
Considerable effort has gone toward trying to understand the antecedents and role of WFC. Research shows individual characteristics and experience influence perception of WFC,6 with two significant implications for the dynamics of WFC: Different individuals may respond to the same WFC differently, and individuals may react to the same WFC differently over time through their attempts to cope with WFC and their changing situations. However, not enough research has considered the dynamics of WFC, especially regarding individual differences. In addition, no one fully understands the relationships between WFC and job and family satisfaction sufficiently due to inconsistent findings about their relationships across different studies.14,17 Moreover, emerging social media could be reshaping the dynamics of WFC and remains unexplored.
Our study began to fill these gaps in the literature by recognizing the great potential of social-media data to help social scientists, as well as business managers, discover the dynamics of WFC and advance understanding of the relationships between WFC and job and family satisfaction in modern society. In addition, the social-media analytics approach we used has significant methodological implications for WFC research. First, work-family (WF) research has been criticized for being overly reliant on cross-sectional designs and self-report survey data at the individual level of analysis.7 In contrast, diary methods are promising for examining WF phenomena over time or to increase the field's understanding of dynamic WF relations.7 Users of social media, particularly Twitter, write brief time-stamped text updates (such as tweets) about their lives on the go, ranging from daily activity to current events, news stories, and their interests in real time. The ability to observe and analyze high-volume, continuous streams of sample data as they are being generated can effectively support the study of WFC dynamics. Second, much WF research has been conducted with a homogeneous population, but the studies "must begin to use large heterogeneous populations."14 Twitter, as a social-media platform, had approximately 320 million active users worldwide per month as of September 2015. The large scale and diversity of the Twitter-user population reflects variations in individual and organization settings that may influence WFC and satisfaction relationships. Third, the most common survey method suffers from recall error due to misinterpretation of survey questions, frailty of memory, close-categories bias, and lack of intrinsic motivation.16 In contrast, self-recorded information about individual lives at the moment it happens in tweets can effectively alleviate recall error. We thus innovatively adopted tweets as the lens to examine WFC in our study.
Work and family are two important domains for all people. WFC is experienced when there is conflict between pressure in either domain.12 WFC can be classified into time- and strain-based categories, along with others.12 Specifically, the time devoted to and the strain produced by work make it difficult to fulfill requirements of family and vice versa. Past research has shown that working adults explicitly experience mood spillover across their work and family lives; the stress in these two areas also increases with the amount of spillover.13 On the other hand, individuals can learn to live with WFC as their experience and response tactics improve, in turn lowering their stress.5 Additionally, family and work situations fluctuate depending on one's circumstances, and a person may reply differently at different times.20 However, the dynamics of WFC were not addressed in previous studies because they generally adopted one-time, cross-sectional measure with few exceptions. For instance, one study published in 2013 collected survey data at two different points in time2004 and 2006to investigate the relationship between WFC and pay satisfaction.1 Although a diary method was used to examine WFC and work-family facilitation,3 finding considerable variation in the same individuals, the traditional diary method is difficult to scale up in terms of number of participants. Further, despite that previous studies established negative correlations between WFC and job and family satisfaction, the strength of relationships varied greatly from one study to another, ranging from nearly negligible to strong.14,17 This variability raises the need to explain the inconsistent findings.
To address these limitations, we propose a social-media analytics approach to investigating the dynamics of WFC, particularly time-based and strain-based WFC, and the relationships between WFC and job and family satisfaction through the lens of tweets for the first time. Twitter provides abundant data where user opinions on certain topics or events can be mined and is expected to present a precise picture of dynamics and influence of WFC and related user experience and perception. Tweets have been used to examine the changing patterns of diurnal and seasonal mood with work, sleep, and length of day,11 but they have yet to be explored to help understand the dynamics of WFC. Our research thus takes a significant step toward expanding research methods for examining WFC.
We used a dataset of Twitter users and their tweets collected through a combination of random sampling and social sub-graph extraction that was representative of the actual population of the U.S.8 To study WFC, we filtered those users who did not have a job based on whether none of their tweets involved work-related topics. We also filtered those users who posted a total of 10 tweets or fewer to assemble a sufficient base for understanding how they experienced WFC. The final sample, as of February 2010, included 3,327,715 tweets from 93,700 users, with 35 tweets per user on average over six months in 2009 and 2010.
To account for time-zone differences among Twitter users, we converted the timestamps on all tweets to users' local time. Our identification of topics in tweets employed Linguistic Inquiry and Word Count, a tool widely used for text analysis,19 providing typical word dictionaries that measure psychological (such as social) processes, and personal concerns. Personal concerns consist of sub-categories (such as work and home), and social processes consist of sub-categories (such as family). We classified tweets with words in the work category as job-related and tweets with words in the family and home sub-categories as family-related.
We used two sets of variables: WFC and satisfaction. We measured all variables first at the individual level by date and then aggregated them over all active users for each date. To account for individual differences, we measured the baselines of all individuals and standardized all measures by removing baseline differences across individuals.
Following previous conceptualizations,4,12 we measured WFC along two dimensions: time and strain. Time-based WFC (TC) is a consequence of competition for an individual's time from work and family responsibilities. One classical example is "The time I spend with my family (work) often causes me not to spend time in activities at work (family)."4 Strain-based WFC (SC) arises when role stressors at work (or with family) induce strain in the individual, hampering fulfillment of role expectations in the family (or work) domain. Two classical examples are "I am often so emotionally drained when I get home from work that it prevents me from contributing to my family" and "Because I am often stressed from family responsibilities, I have a hard time concentrating on my work."4
TC. We used the proportion of tweets on work- and family-related topics posted by a user on a given day as a proxy for the time the user reported spending in the respective domain. TC arose when an individual allocated above-average time for work and simultaneously below-average time for family and vice versa. Accordingly, we defined TC as the average difference between the time allocation for work and for family.
SC. Strain can be deduced from psychological and physical dimensions,12 though it is difficult, if not impossible, to measure physical strain directly from social-media data. We thus measured SC from the psychological aspect of participant data, specifically through negative mood (such as anxiety, anger, and sadness). Negative mood in one domain is associated with pessimism and rumination, causing individuals to neglect requirements in the other domain.18 SC arose when an individual simultaneously experienced above-average negative mood in both work and family. Accordingly, we defined SC as the average negative mood across work and family.
We measured satisfaction in two domainswork and familywe labeled as job satisfaction (JS) and family satisfaction (FS), respectively. Satisfaction can be explained as a causal sequence linking mood to performance and reward in a domain.18 Positive mood can facilitate role performance by enhancing cognitive functioning, increasing task activity and persistence, and promoting positive interactions with others. Meanwhile, intrinsic and extrinsic rewards earned through role performance can thus enhance positive mood.18 We viewed positive mood as a proxy measure of satisfaction.
JS (or FS). We defined satisfaction with work (or family) as the proportion of work- or family-related tweets expressed in positive moods.
To understand the dynamics of WFC, we first analyzed the trends of time-based WFC (TC) and strain-based WFC (SC) at different levels of time granularity after removing the data associated with American national holidays (such as Thanksgiving, Christmas, and New Year's). The results show the trends of WFC at levels of month and week are both relatively stable. Nevertheless, WFC fluctuates by day of the week (see Figure 1).
Figure 1a shows TC is heightened during weekdays relative to weekends. This observation confirms previous findings on work-family time allocation that both men and women in the U.S. spend more time in domestic work and caring for children on a weekend day than on a weekday.10 In addition, TC shows an increasing trend from Sundays to Tuesdays and the opposite trend from Thursdays to Saturdays, peaking on Tuesdays and dipping on Saturdays. There is thus a sharp increase in TC when transitioning from weekends to weekdays and a sudden drop during the opposite transition. Note TC stays in the negative range, indicating family-to-work spillover events are more prevalent than work-to-family spillover.
Figure 1b shows SC is much greater on weekdays than weekends, except for Wednesdays. If the trend profile of TC resembles a normal distribution, then the trend profile of SC simulates a bimodal distribution, with Mondays and Thursdays being two peaks. Unlike TC, which peaks on Tuesdays, SC reaches a peak immediately after weekdays begin. The abrupt drop in SC on Wednesdays could be explained by people reaching a state of work-family balance by the middle of the week. A sharply elevated SC on Thursdays may be attributed to the pressure of trying to complete scheduled weekly tasks, onset of work exhaustion, or preparation for upcoming family commitments.
Overall, the results show both TC and SC vary markedly with the day of the week, an observation that helps explain the inconsistent findings about the form and intensity of WFC in the literature. They also imply a major limitation of survey-based methods that have dominated traditional WFC research. Depending on the day a survey is administered, participants' responses can vary greatly.
Figure 2 shows the trends for job satisfaction (JS) and family satisfaction (FS), including that individuals' JS and FS are both higher on weekends than on weekdays. People feel satisfied when their criteria are met, and satisfaction can be viewed as a degree of realized expectation. On weekdays, the expectation of work performance is high, and, consequently, the possibility of unfulfilled expectation may also be high, negatively influencing JS. In comparison, the expectation is lower on weekend days, indirectly contributing to a higher level of JS.
JS reaches its highest level on Fridays, as in Figure 2a, a finding that could be attributed to actual fulfillment of weekly work expectations. At the other end, JS is at its lowest on Tuesdays, followed by an increasing trend the rest of the week. One possible explanation is work stress peaks on Tuesdays when the weekend-to-weekday transition is over and work expectations begin to mount.
FS peaks on Wednesdays, as in Figure 2b, a finding that could be attributed to people's state of work-family balance, as manifested in both low SC and relatively low TC, as in Figure 1. There is a sharp drop in FS from Sundays through Tuesdays, as expected. Surprisingly, FS barely makes it into the positive range on Saturdays. One possible explanation is that employed women shift child care and housework to weekend days.9 Given individuals' high expectation of family commitments on weekends, it is difficult for them to feel highly satisfied, particularly as weekends begin to unfold. In comparison, FS is relatively more stable than JS, as reflected in a smaller variance across different days, as in Figure 2a and Figure 2b.
To understand the association between WFC (TC and SC) and satisfaction (JS and FS) we performed pairwise correlation analyses between the two sets of variables. The results reported in the accompanying table reflect that all correlation coefficients are negative. In addition, the correlations between SC and JS/FS, but not between TC and JS/FS, are significant (p<.001).
The findings on the relationship between TC and satisfaction challenge current mainstream thinking that WFC has a negative influence on job and life satisfaction.15 Employees could adapt themselves to WFC as their experience and responding tactics improve through empowerment of modern communication technologies. For instance, telecommunicating could support employees in performing some or all of their job functions outside the workplace, even as they stay connected to work during non-work hours.2 As the boundary between family and work activities blurs in some situations, and despite the distinct norms and requirements of the two roles,12 TC inevitably loses its influence on JS and FS.
Our findings on SC emphasize the importance of employees' psychological well being, with significant managerial implications for human resource management in organizations. Minimizing employees' distress, anxiety, fear, anger, and disgust is thus instrumental to boosting their JS and FS. For example, an increasing spillover between work and family activities might contribute to both increased SC and decreased JS and FS. Managers looking to control employees' stress try to avoid assigning work activities for non-work hours. And, while enjoying a flexible work-family arrangement, employees are able to reduce their psychological strain by minimizing interference of family responsibilities in non-family situations and non-family hours.
These results also show that both JS and FS are subject to the influence of the same type of WFC, highlighting the importance of WFC in our lives and significant spillover between our personal experience of work and family.
WFC garners widespread attention in modern society beyond human resources management. Despite extensive research in this area, different studies report inconsistent and even contradictory findings on the effects and intensity of WFC. Additionally, the overlap in time and place between traditional family and work roles may also introduce new opportunities for WFC to manifest in people's everyday lives.
Ours is the first study to investigate the dynamics of WFC and explain mixed findings concerning the relationships between WFC and satisfaction in the literature. Our results show WFC is markedly higher on weekdays than weekends and, more important, fluctuates on weekdays. Our comparison of the two types of WFC shows the dynamics of strain-based conflict is more pronounced than time-based conflict on weekdays. Interestingly, we found strain-based conflict reaches its lowest point on Wednesdays. We also found the relationships between WFC and job and family satisfaction, suggesting that, while people may adapt to the inherent conflict between work and family activity due to the flexibility of work place and time, they also feel dissatisfaction from connecting with work during non-work hours.
The social-media analytics we employed address the limitations of survey methods dominating traditional WFC studies. High-volume, high-velocity Twitter data provides a dynamic, fine-grain view of individuals' behavior in a naturally occurring setting that serves as an ideal testbed for understanding WFC.
This research can be improved and continue in several directions. Tracking a larger number of Twitter users over a longer period of time would improve the general applicability of the findings. In addition, some jobs have distinctive busy and off-peak seasons. In view of country differences in work-family time10 and workweek,11 WFC outside the U.S. deserves its own separate investigations. Alternative techniques should be explored to improve extracting work- and family-related topics and mood from social-media data. And resolving multiple online identities for the same Twitter users can help refine the findings of our study.
We thank the reviewers for their constructive comments. Portions of this work are supported by the National Science Foundation under Award Number SES-152768. The research reported here is that of the authors and does not reflect the official policy of the National Science Foundation.
3. Butler, A.B., Grzywacz, J.G., Bass, B.L., and Linney, K.D. Extending the demands-control model: A daily diary study of job characteristics, work-family conflict, and work-family facilitation. Journal of Occupational and Organizational Psychology 78, 2 (June 2005), 155169.
6. Carr, J.C., Boyar, S.L., and Gregory, B.T. The moderating effect of work-family centrality on work-family conflict, organizational attitudes, and turnover behavior. Journal of Management 34, 2 (Apr. 2008), 244262.
8. Cheng, Z., Caverlee, J., and Lee, K. You are where you tweet: A content-based approach to geo-locating Twitter users. In Proceedings of the 19th ACM International Conference on Information and Knowledge Management (Toronto, ON, Canada, Oct. 2630). ACM Press, New York, 2010, 759768.
14. Kossek, E.E. and Ozeki, C. Work-family conflict, policies, and the job-life satisfaction relationship: A review and directions for organizational behavior-human resources research. Journal of Applied Psychology 83, 2 (Apr. 1998), 139149.
17. Michel, J.S., Kotrba, L.M., Mitchelson, J.K., Clark, M.A., and Baltes, B.B. Antecedents of workfamily conflict: A meta-analytic review. Journal of Organizational Behavior 32, 5 (July 2011), 689725.
20. Taylor, B.L., DelCampo, R.G., and Blancero, D.M. Workfamily conflict/facilitation and the role of workplace supports for U.S. Hispanic professionals. Journal of Organizational Behavior 30, 5 (July 2009), 643664.
21. U.S. Census Bureau. America's Families and Living Arrangements: 2015: Adults (A Table Series); https://www.census.gov/hhes/families/data/cps2015A.html
©2017 ACM 0001-0782/17/06
Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and full citation on the first page. Copyright for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or fee. Request permission to publish from email@example.com or fax (212) 869-0481.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2017 ACM, Inc.