the paradox of family structure and plans after work: why single … · 2019. 4. 20. · between...

r Academy of Management Journal2018, Vol. 61, No. 4, 1231–1252.https://doi.org/10.5465/amj.2016.0086

THE PARADOXOF FAMILY STRUCTURE AND PLANSAFTERWORK: WHY SINGLE CHILDLESS EMPLOYEES MAY BE THE

LEAST ABSORBED AT WORK

TRACY L. DUMASThe Ohio State University

JILL E. PERRY-SMITHEmory University

Existing research has shown that positive family experiences can affect work positively.In this article, however, we consider how family may enhance work even when familyexperiences are not explicitly positive. We draw on boundary theory and cognitivepsychology’s current concerns theory to evaluate how employees’ family structures andassociated after-work activities affect their work absorption. A survey of business schoolalumni (Study 1) revealed that single, childless workers reported lower absorption thanworkers with other family structures. Further, a daily diary study of university em-ployees (Study 2) showed that employees’ planned after-work activities explained therelationship between family structure and work absorption. Specifically, single, child-less workers anticipated fewer domestic after-work activities, resulting in lower workabsorption. Due to similarities between domestic responsibilities and work tasks—forexample, their obligatory and goal-directed nature—anticipating domestic responsi-bilities after work reinforces, rather than distracts from, the work mindset, thus keepingemployees more immersed psychologically in their work. This finding suggests thathaving a spouse and/or children can affect employees’ work absorption positivelythrough the anticipation of domestic duties after work. Thus, our study contributes toa more comprehensive view of how employees’ work and nonwork lives are connected.

“Janet’s perfect for that job. Because for that job, youhave to have no life. Janet has no family. Perfect.”

—Former Pennsylvania governor EdwardRendell regarding former Arizona governor

Janet Napolitano’s nomination for Secretaryof Homeland Security (Collins, 2008)

As illustrated in the opening quote, familystructure—the presence or absence of a spouse and/

or children—is often treated as a signal of an in-dividual’s ability to devote themselves to work, dueto assumptions about how family structure predictslife outside of work. Organizational research docu-ments these assumptions extensively. For example,within Wall Street financial services firms, assump-tions about workers’ family responsibilities are re-lated to differences in pay and advancement (Roth,2003). Moreover, supervisors’ perceptions that em-ployee family responsibilities compete with workhave a negative impact on employee promotions andperformance appraisals, regardless of whether theemployees actually face difficulty managing workand family role responsibilities (Hoobler, Wayne, &Lemmon, 2009). These traditional expectations arereified in scholarship on the “ideal worker,” a termrecognizing the historical organizational preferencefor “unencumbered” workers who can devote theirenergyandattentionprimarily to thework role (Bailyn,Drago, & Kochan, 2001; Reid, 2015; Williams, 2001).

Although the original “ideal worker” conceptevoked a specific family structure—men with wiveswho were not employed (Williams, 2001)—more

We thank our editor Carol Kulik for her gracious andinvaluable guidance on this manuscript. Our three anon-ymous reviewers also contributed thoughtful and usefulfeedback. We thank Nora Madjar, Racheli Perl, AshleighRosette, and Nancy Rothbard for their helpful commentsoffered in the early stages of this project, and we thankKristopher Preacher for directing us to multilevel struc-tural equation modeling analysis. Also, seminar partici-pants at Emory University, London Business School, theOhio State University, Syracuse University, and Univer-sity College London provided helpful suggestions. Earlierversions of this paper were presented at the Academy ofManagement conference and the Work and Family Re-searchers Network conference.

1231

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download, or email articles for individual use only.

https://doi.org/10.5465/amj.2016.0086

recently, single, childless workers have been de-scribed as fitting the ideal worker model, due toexpectations that they have few domestic re-sponsibilities to detract from their focus on work(Blair-Loy, 2004). Yet, whereas research documentsthevery real challenges ofmanagingwork and familyresponsibilities (see Eby, Casper, Lockwood,Bordeaux, & Brinley, 2005, for a review), there islittle evidence that family detracts consistently fromemployees’ work, and limited research comparingthose with nuclear families to those without. Thisomission is problematic, given that single, childlessworkers are a growing segment of the workforce(Hamilton, Gordon, & Whelan-Berry, 2006), cur-rently comprising nearly 36% of all workers(U.S. Bureau of Labor Statistics, 2014). In this paper,we focus on single, childless employees to examinethe connections among employee family structure,after-work activities, and “work absorption”—acentral component of work engagement defined asa flow-like state of full immersion (Rodriguez-Sanchez, Schaufeli, Salanova, Cifre, & Sonnen-schein, 2011; Rothbard, 2001; Schaufeli, Bakker, &Salanova, 2006), critical for complex work (Hobfoll,2011). Building on boundary theory (Ashforth,Kreiner, & Fugate, 2000), Richter’s (1990) conceptof anticipatory boundary transitions, and currentconcerns theory (Klinger, Barta, & Maxeiner, 1980),we posit that employees’ anticipated after-work activ-ities, drivenby family structure, affectwork absorptionin a manner contrary to traditional expectations.

Whereas early work–family studies fell within thedepletion perspective (Rothbard, 2001), primarilyaddressing the negative effects of family on work,such as interrole conflict (Coverman, 1989; Frone,Russell, & Cooper, 1992; Greenhaus &Beutell, 1985),more recent scholarship increasingly addresses theways that employees’ nonwork lives, includingfamily andother roles, canhave anet positive impacton their work (Dumas & Stanko, 2017; Greenhaus &Powell, 2006; Rothbard, 2001). Work–family “en-richment” scholarship considers mechanisms suchas skill transfer, social support, and positive affectthrough which having a family can impact work ex-periences positively (e.g., Dumas & Stanko, 2017;Graves, Ohlott, & Ruderman, 2007; Greenhaus &Powell, 2006), even while coexisting with negativeeffects of family responsibilities (Rothbard, 2001;Ten Brummelhuis, van der Lippe, & Kluwer, 2010).Similarly, research on prosocial motivation showsthat the need to support one’s family can serve asa source of work motivation, energizing employeeeffort and improving performance (Menges, Tussing,

Wihler, & Grant, 2017). Also, research based on theeffort–recovery model (Meijman & Mulder, 1998)shows how after-work activities can replenish em-ployees’ energy and cognitive resources, rechargingtheir return to work the next day (Sonnentag, 2003;van Hooff, Geurts, Beckers, & Kompier, 2011).

With the growing bodyof enrichment research, thequestion arises as to whether family or nonwork ex-periences are only beneficial for work when they areexplicitly positive. Also, whereas the motivation tosupport one’s family can increase employee moti-vation on simple work tasks that are not intrinsicallymotivating (e.g., Menges et al., 2017), it is less clearhow family responsibilities affect work contribu-tions in jobs that are more complex and inherentlymore engaging (Mainemelis, 2001; Quinn, 2005).Wepropose that these issues may be addressed, in part,by examining the connection between family struc-ture and activities employees anticipate at the end ofeach workday. An existing body of research focuseson how after-work activities affect employees’ re-covery the following workday (Oerlemans, Bakker,& Demerouti, 2014; Sonnentag & Zijlstra, 2006), butwe flip this focus to highlight how anticipatingafter-work activities affects absorption during theworkday.

Our perspective contributes to existing literature inseveral ways. We contribute to theories addressingthe positive impact of family on work by consideringa mechanism other than the transfer of affect or skills(e.g., Greenhaus & Powell, 2006; Rothbard, 2001) orthe expansion of available resources through moti-vation (e.g., Menges et al., 2017). Rather, we focus onthe psychological effects of anticipating future activ-ities whereby anticipating goal-directed, obligatoryactivities helps induce a mindset conducive to ab-sorption at work. This suggests that having a familycan affect work absorption positively whether or notthe familyexperience ispositive, andoffers adifferentperspective on the effects of after-work activities. Inaddition, we contribute by exploring the connectionbetween family structure and nonwork activities.Assumptions abound regarding the effects of familystructure on employees’ activities outside of work(SHRM Online Staff, 2011); however, in this paper,weassess this issuedirectly.Accordingly,webroadenthe consideration of the work–family interface byhighlighting single, childless workers who remainunder-examined and largely absent from existingwork–family studies (Wilson & Baumann, 2015, isa notable exception). With this research, we seek tocontribute to the development of a more compre-hensive view of the work–nonwork connection.

1232 AugustAcademy of Management Journal

FAMILY STRUCTURE, AFTER-WORKACTIVITIES, AND WORK ABSORPTION

We examine the connection between family struc-ture and employees’ “work absorption”—the inten-sityofconcentration, focus,andpsychological immersionwhile working (Kahn, 1990; Rothbard, 2001).Absorption addresses a central and uniquely psy-chological aspect of work engagement (Kahn, 1990;Rodriguez-Sanchez et al., 2011; Rothbard, 2001;Rothbard & Patil, 2012; Schaufeli & Bakker, 2004).When employees are absorbed, they feel lost in timeand are fully present psychologically—an outcomeparticularly relevant for knowledge workers whomust apply cognitive resources to solve complexproblems (Quinn, 2005). Therefore, we consider theconnection between family structure and work ab-sorption in the context of complex, knowledge-basedwork for which absorption is critical, but alsowhere the work itself can be interesting and com-pelling (Mainemelis, 2001; Quinn, 2005). Thus, weaddress processes distinct from those covered inresearch addressing the effects of family-relatedmotivation on low-skilledwork,whereby the desireto help one’s family increases work effort, but onlywhen the work provides little intrinsic motivation(e.g., Menges et al., 2017).

Research on boundary theory (Ashforth et al.,2000), which addresses the nature of the work–nonwork interface and employee transitions be-tween work and nonwork domains, provides insightinto how employees’ work absorption may relateto their family structure and after-work activities.According to boundary theory, employees oftentranscend the work–nonwork boundary cognitivelyby thinking about nonwork activities while they arestill at work (Dumas & Sanchez-Burks, 2015; Glavin,Schieman, & Reid, 2011; Nippert-Eng, 1996). In es-sence, they are making mental transitions to theirnonwork roles before the workday is over (Ashforthet al., 2000; Smit, Maloney, Maertz, & Montag-Smit,2016). This is particularly the case for nonwork rolesthat are highly salient to employees, such as beinga spouse or parent (Ashforth et al., 2000; Dumas &Stanko, 2017). Boundary research also suggests thatdifferences in the types of after-work activities mayshape employees’ experiences during the workday.For example, in a qualitative study of daily work-to-home transitions among middle managers in dual-income couples, Richter (1990: 145) identifiedseveral transition styles, including the “anticipa-tory” transition style whereby the “psychologicaltransition precedes the physical transition,” and

observed that thenature of the anticipated after-workactivity affectedmanagers’mindsets at the end of theworkday. Interviewees who anticipated going hometo housework or childcare duties worked harder to-ward the end of the workday and remainedmentallypreoccupied with work even on the commute home,whereas those anticipating relaxation or recreationafter work detached psychologically from work ear-lier in the day. Unlike our focus on family structureand work absorption, Richter (1990) addressed gen-der differences in work–home transitions. However,the pattern she observed suggests that anticipatedactivities affect mental transitions away from workand thus affect work absorption.

Anticipating Domestic and Leisure Activitiesafter Work

Scholars have studied a variety of after-work ac-tivities, and domestic and leisure activities figureprominently in existing research (Oerlemans et al.,2014; Saxbe, Repetti, & Graesch, 2011; Sonnentag &Zijlstra, 2006). Drawing on this research, we define“domestic activities” as those associated with main-taining a household, including cooking, cleaning,laundry, home repairs, managing finances, andchildcare duties (e.g., bathing, feeding) (Oerlemanset al., 2014; Vernon, 2010). We define “leisure ac-tivities” as those pursued for recreation, includingsocializing, sports, outings to cultural events, orother forms of entertainment (Lee, Lawson, Chang,Po-Ju, Neuendorf, & Dmitrieva, 2015; Sonnentag &Grant, 2012). In general, domestic activities areobligatory, goal directed—pursued for the sakeof achieving an objective—and require effort (Leeet al., 2015; Ten Brummelhuis & Bakker, 2012).Conversely, leisure activities are voluntary, char-acterized by freedom, and are intrinsically moti-vating, with the goal of relaxation or enjoyment(Csikszentmihalyi & LeFevre, 1989; Iso-Ahola, 1997;Zijlstra & Cropley, 2006).

The effects of anticipating domestic or leisure ac-tivities onwork absorption can be explained, in part,by current concerns theory (Klinger et al., 1980),which addresses factors inducing one’s mind wan-dering away from the current task. According to thistheory, people focus on stimuli that are the mostcompelling (Smallwood, 2013). When future eventsare novel, nonroutine, or voluntary, they may beconsidered more compelling than current tasks(Forster & Lavie, 2014; Haas, Criscuolo, & George,2015). For example, when an employee is at work(current task), they may anticipate obligatory,

2018 1233Dumas and Perry-Smith

routine domestic tasks—laundry or a sink full ofdirty dishes. This anticipation may render themabsorbed at work because it reinforces a goal-directed mindset, but also because the anticipatedtask may be less compelling than work (the currenttask), particularly if work is inherently engaging.However, anticipating a concert with friends afterwork may make it more difficult for the employee toconcentrate at work, because the novel, nonroutine,voluntary anticipated activity takes the employeeout of a goal-directedmindset, andmay also bemoresalient or compelling than work (the current task).

Our characterization of the differential contrastof domestic and leisure activities with work alsodraws from research addressing the impact of after-work activities on employee recovery from work(Sonnentag & Zijlstra, 2006). A critical driver of therecovery process, whereby people replenish the re-sources expended through work effort, is “psy-chological detachment” from work, described ascompletely disengaging from work, refraining fromthinking about work (Etzion, Eden, & Lapidot, 1998),or mentally “switching off” fromwork (Sonnentag &Fritz, 2007). Psychological detachment from work isgreater to the extent that after-work activities presenta notable departure from work tasks, or draw ona different set of resources than the individual’s re-curring pattern of work activities (Etzion et al., 1998;Sonnentag & Fritz, 2007). Leisure activities presenta greater departure from work relative to domesticactivities because domestic activities draw on someof the same resources as work tasks (Lee et al., 2015;Oerlemans et al., 2014; Saxbe et al., 2011; Sonnentag& Braun, 2013; Sonnentag & Zijlstra, 2006). In re-covery research, greater detachment resulting fromafter-work leisure activities can increase engage-ment on the subsequent workday (Sonnentag &Braun, 2013). However, we expect that anticipatingleisure activities can reduce work absorption duringthe current workday. Conversely, since domesticactivities are less likely to induce psychological de-tachment afterwork (Sonnentag&Zijlstra, 2006), butrather may heighten goal direction and a workmindset, we expect that anticipating them may in-crease work absorption.

Importantly, our arguments differ from the com-pensation model (Edwards & Rothbard, 2000), whichexplains that employees may invest more in theirwork to make up for negative family experiences.Rather than focusing on negative or deficient experi-ences in the work and nonwork domains, we addressthe extent to which after-work activities are goal di-rected and obligatory, as are work tasks, or present

a departure from the work routine. Further, we focuson absorption, which is an emergent state (Rodriguez-Sanchez et al., 2011; Rothbard, 2001) as opposed toa conscious choice, such as staying at work late toavoid going home (e.g., Hochschild, 1997).

Family Structure and After-Work Activities

The expectation that family structure predictsafter-work activities is embedded within our exam-ination of family structure, after-work activities, andwork absorption. People devote more time to rolesthat are important to their identities or self-concepts(Ashforth&Mael, 1989; Stryker, 1968)—and spousesor parents are more likely to view the family roleas central to their identities than are single, child-less employees (Dumas & Stanko, 2017; Wilson &Baumann, 2015). Naturally, parents spend time at-tending to their children’s needs (Aryee & Luk,1996). Moreover, time-use scholars have estab-lished that married couples spend more time thansingles on responsibilities such as cleaning,whether or not they have children. Since marriedcouples have partners for sharing household tasks,they choose to perform those tasks themselves ratherthan outsource them, yet, evenwith task sharing, theystill devote more time to domestic activities than dosingle, childless people (Vernon, 2010). For example,married couples—whether parents or not—preparemore of their meals at home than single, childless in-dividuals do (Kroshus, 2008). Also, married couplestypically live in larger residences than singles do, andthey are more likely to be homeowners and thereforespend more time on home maintenance (Vernon,2010). We thus expect single, childless employees toanticipate fewer domestic activities after work thanemployees with other family structures.

Conversely,we expect single, childless employeesto anticipate more leisure activities after work thanthose with other family structures. Single, childlessemployees are more likely than employees withother family structures to identify with their per-sonal, non-family roles (Wilson & Baumann, 2015).Thus, we expect that they are more likely to devotetheir time to leisure pursuits than someone witha family would. Similarly, due to the high valuemodern society places on workplace achievementand family devotion, workers who are married and/or have children often sacrifice leisure time to spendmore time either at work or with their families(Hochschild, 1997; Nomaguchi & Bianchi, 2004;Schor, 1991). Importantly, some scholars note thatthis discrepancy in leisure activities between single,


childless employees and those with other familystructuresmaydependon the type of leisure activitiesconsidered (Lee & Bhargava, 2004). For example,parents devote time to family-based leisure activitiessuch as taking children to the zoo (Bittman &Wajcman, 2000). However, in general, existing find-ings support the overall argument that single, child-less employees pursue more leisure activities in theirnonwork time than do employees with families.

We argue that differences in after-work activitieshelp explain the relationship between family structureand work absorption. Since single, childless workerstend to anticipate fewer domestic responsibilities andmore leisure activities afterwork compared toworkerswith other family structures, we expect that single,childless employees will report lower work absorp-tion. We also expect that the anticipation of domesticactivities and leisure activities after work will mediatethis effect. We therefore predict the following:

Hypothesis 1. There is a significant relationship be-tween family structureandworkabsorption, such thatsingle, childless employees report lower work ab-sorption than employeeswith other family structures.

Hypothesis 2. There is a positive relationship betweendomestic activities and work absorption, such thatanticipating domestic activities after work increaseswork absorption.

Hypothesis 3. There is a negative relationship be-tween leisure activities and work absorption, suchthat anticipating leisure activities after work de-creases work absorption.

Hypothesis 4. There is a significant indirect effectof family structure on work absorption operatingthrough the anticipation of domestic activities afterwork, such that, compared to workers with families,single, childless people anticipate fewer domesticactivities after work—and anticipating fewer domes-tic activities results in lower work absorption.

Hypothesis 5. There is a significant indirect effectof family structure on work absorption operatingthrough the anticipation of leisure activities afterwork such that, compared to workers with families,single childless workers anticipate more leisure ac-tivities after work—and anticipating more leisureactivities results in lower work absorption.

STUDY 1

Sample and Procedure

We randomly sampled 1,966 undergraduate busi-ness and master of business administration (MBA)

alumni of a private U.S. university as part of a largerwork–life study. Sample members had graduatedfrom between 69 years to 1 year prior to data col-lection. This sample allowed for a variety of familystructures, and business graduates tend to assumecomplex and skilled jobs for which absorption ishighly relevant (Hobfoll, 2011; Quinn, 2005).

We used a mixed-mode, four-contact strategy ofsending a web-based survey via email first, followedby a combination of email and paper reminders(Dillman, 2000). We received a total of 562 surveys(for a 29% response rate). We removed respondentswho were not currently employed (n5 46) or retired(n 5 47) from the sample because they could notanswer questions about their current work absorp-tion. This resulted in 469 available respondents, ofwhom 61.8% were male. Their average age was39.07, and 82.4% were Caucasian. Organizationaltenure was as follows: 42.2%, less than 3 years;34.7%, 3 to 10 years; 23.1%,more than 10 years; and75.7% of respondents held a master’s degree orhigher. Married respondents made up 72.9% of thesample, and the mean number of children for all re-spondents was .94. Finally, 81.7% reported occu-pations in management and business functions,based on the U.S. Department of Labor Bureau ofLabor Statistics standard occupational classifica-tions. Other professions (18.3%) included lawyers,realtors, architects, and engineers.

The final sample size was 353 (75.3% of the 469available participants), due to listwise deletion formissing data.1 The final sample used in the analysisis similar to all available respondents in terms ofgender distribution (x2(1) 5 0.23, p . .05), maritalstatus (x2(1) 5 0.01, p . .05), level within the firm(x2(3)5 0.08, p. .05), highest degree (x2(2)5 0.20,p. .05), andwork absorption (x2(5)5 2.18,p. .05).

Measures

Work absorption. We measured work absorptionusing four items from the absorption subscale ofRothbard’s (2001) engagement scale. The work ab-sorption construct captures the extent to which in-dividuals become engrossed in their work, and it iscommon across engagement measures used by dif-ferent scholars (Rothbard & Patil, 2012; Schaufeli& Bakker, 2004; Viljevac, Cooper-Thomas, & Saks,

1 As this study was part of a large-scale work–life project,therewas attrition as the surveyprogressed. Family structureand demographic controls—primary contributors to the de-crease in sample size—were at the end of the long survey.


2012).Weaveraged the following four items to reflectwork absorption: “When I am working, I often losetrack of time”; “I often get carried away by what I amworking on”; “Nothing can distract me when I amworking”; “When I amworking, I am totally absorbedby it” (rated from 1, “strongly agree,” to 7, “stronglydisagree”) (a 5 .78).

Family structure. We coded family structure tocompare singles without children to all other familystructures. The singles category included never-married (86.7%), divorced (10%), and widowed re-spondents (3.3%). The married category includedmarried (98.8%) and same-sex domestic partnership(1.2%) respondents.2 We coded respondents as par-ents if they had minor children in the home, con-sistent with research on similar populations (Graveset al., 2007; Hammer, Neal, Newsom, Brockwood, &Colton, 2005). We coded this variable 1 for single/childless and 0 for all other family structures.

Control variables. Gender is central to work–lifemodels (Eby et al., 2005), and it is also a major com-ponent of the ideal worker norm (Kelly, Ammons,Chermack, &Moen, 2010; Reid, 2015). Therefore, wecontrolled for the respondents’ sex as a proxy forgender (0 5 male and 1 5 female).

Older respondents may be more skilled in man-agingmultiple role demands, and agemay influencethe priority of work over nonwork (Gordon &Whelan, 1998; Martins, Eddleston, & Veiga, 2002).Moreover, contextual factors may have a greater ef-fect on engagement among older workers (Avery,McKay, & Wilson, 2007). Therefore, we controlledfor age, measured in years.

Last, we controlled for job level, since higher po-sitions provide more complexity, autonomy, andsignificance, which enhance absorption (Rothbard &Patil, 2012). Respondents indicated their level inthe organization (1 5 entry level, 2 5 lower level,3 5 middle level, 4 5 senior level).

Results and Discussion

Table 1 presents the means, standard deviations,and correlations. Hypothesis 1 predicted that single,childless respondents would report lower levels ofabsorption. We tested this hypothesis via ordinaryleast squares regression.Aswe expected, the effect of

being single and childless on work absorption wasnegative and significant (b 5 2.31, SE 5 0.13, p ,.05). The full equation is:

absorption5 4:062 :073 gender2 :0023 age

1 :133 joblevel2 :313 single=childlessðn5 353Þ

Hypothesis 1 was thus supported.Severalquestions remain,however.WithourStudy

1 data, we were not able to test whether anticipatingafter-work activities mediates the effect of familystructure on absorption. Additionally, the Study 1data consisted of perceptions collected from a singlesurvey instrument, which limits the causal inferenceof the study and can result in common method vari-ance (Podsakoff, MacKenzie, & Podsakoff, 2012).

STUDY 2

We designed Study 2 to test the indirect effect offamily structure on absorption through its effects onemployees’ anticipation of domestic or leisure ac-tivities after work. We also sought to examine howwork absorption might change from day to day asa result of employees’ nonwork activities, given priorstudies showing daily fluctuations in engagement(e.g., Sonnentag, 2003). Therefore, in Study 2, we uti-lized a daily diary surveymethodology extending overfive workdays, as well as a one-time general survey.

Sample and Procedure

We surveyed employees of one nonacademic di-vision of a large, private university. The division wascomposed primarily of professional staff engaged inone area of administration, reporting to the uni-versity’s leadership team. As with Study 1, we useda four-contact strategy (Dillman, 2000), consisting ofan introduction to the study and a series of reminders.

Respondents completed the general survey at leastthree weeks before the first daily diary survey. Thegeneral survey measured all person-level variables,such as marital and parental status, and some of thecontrol variables. Out of 216 invited employees, 103participated in the general survey (for a response rateof 48%). Salaried professional staff were 88.4% ofthe sample, and the remaining 11.6% were hourlysupport staff. Organizational tenure was as follows:28.4%, less than1year; 25.3%,1 to3years; 36.8%,3 to10 years; and 9.5%, over 10 years. Of the participants,46.8%weremarried, and themeannumberof childrenamong all respondents was 1.75. Participants’ agesranged from25 to 65 (M5 43.35, SD5 10.81).Women

2 There was no separate category for heterosexual cou-pleswho cohabitate. These respondents likely chose either“married” or “single,” depending on their view of theirrelationship, or left this question blank. Thus, this is aconservative test.


made up 81% of the sample; Caucasian respondentsmade up 84% and African Americans 14%.

The daily diary survey measured all within-person, day-level variables. For five consecutivedays, participants received two surveys, one at thebeginning of their day that featured questions abouttheir anticipatedplans afterwork, andone at 3:00PMwith work absorption items. Given that participantshad varying work stop times, we chose 3:00 PM forthe afternoon survey as it was far enough into the dayto assess absorption throughout the day, yet earlyenough to precede the day’s end. Out of the 103general survey participants, 97 completed at leastone daily survey.We received 354matchedmorningand afternoon responses (73% response rate). Thisrate represents a mean of 3.6 matched responses perparticipant, suggesting broad participation acrossrespondents. The person-level and day-level samplesizes are consistent with other daily diary studies(e.g., Ilies, Keeney, & Scott, 2011; Sonnentag &Zijlstra, 2006).

Measures

Independent and dependent variables. We mea-sured work absorption using the same items usedin Study 1. The items were reworded, however, toreflect that specific day (e.g., “Today, nothing coulddistract me when I was working”) (a 5 .99). Thefamily structurevariablewas identical to thatofStudy1, in which 1 5 single, childless (n 5 23) and 05 allothers (n5 66).

Proportion of domestic and proportion of leisureactivities.On the dailymorning survey, participantswere asked, “Do you have any nonwork related ac-tivities, commitments, or responsibilities plannedfor the period of time after your regular workday?”Participants who answered “yes” were asked to listup to five activities (mean 5 1.3) and to categorizeeach activity from a list of four options: domestic or

family-related responsibility, leisure, community/volunteer commitment, or personal developmentand wellness. Examples of activities participantscategorized as domestic included going to the gro-cery store, cooking dinner, taking children to or fromactivities, and pet care. Examples of leisure activitieslisted included going out for dinner with friends,family outings, or attending cultural or sportingevents. Respondents reported an average of 0.78planneddomestic and0.24planned leisure activitiesacross the five days. Appendix A includes moreexamples of activities categorized as domestic orleisure.

To calculate the respondents’ daily anticipatedproportion of domestic and proportion of leisureactivities, we divided the number of domestic ac-tivities and the number of leisure activities each bythe total number of activities listed for the day.3

Thus, the proportion values for each ranged from0 to1.On somedays, therewere participantswhodid nothave plans. In this case, the proportion was set to 0.Using the proportion variables is appropriate be-cause they take into account the other types of an-ticipated activities and control for the total numberof activities listed, thus capturing the overall, domi-nant character of the activities.

Control variables. Importantly, although Study 2consisted primarily of daily surveys collectingwithin-person data (i.e., daily after-work activities

TABLE 1Means, Standard Deviations, and Correlations in Study 1

Variable Mean SD 1 2 3 4

Work absorption 4.31 1.00Gender 0.39 0.49 2.07Age 38.52 10.30 .07 2.23**Job level 3.29 0.77 .13** 2.20** .48**Single childless 0.24 0.43 2.16** .16** 2.28** 2.29**

n 5 353.**p , .01

3 In an additional analysis, we analyzed the effects ofeach category of anticipated activities. Although single,childless employees reported both more personal devel-opment and community volunteering activities after work,neither of these explained the effect of family structure onabsorption. In particular, neither had significant effects onabsorption, and neither the indirect effects of anticipatingpersonal development nor community activities weresignificant. This result suggests that the relevant distinc-tion is between domestic and leisure activities.


and daily work absorption), our independent vari-able, family structure, is a between-person variable.Therefore, includingbothwithin- andbetween-personcontrol variables is appropriate for our full model.Similar toour job levelcontrol inStudy1,wecontrolledfor job differences specific to the sample, reflectingdifferent levelsof jobcomplexity thatwouldpotentiallyaffect absorption. An organization informant suggestedthat the least ambiguous indicator of job complexityand schedule variation was monthly pay, since non-exempt workers with fixed hours were paid biweekly(1) and exempt workers were paid monthly (2). As inStudy 1, we controlled for age and sex.

Additional controls were relevant to our daily di-ary design, although not relevant to the cross-sectional design used in Study 1. General trait-likeabsorption may affect employees’ daily absorption,so, following prior diary studies (e.g., Sonnentag,2003), we controlled for respondents’ overall ab-sorption, collected on the general survey, as mea-sured in Study 1 (a5 .72). Last, we controlled for thestart time of the earliest after-work activity, as thisallowed us to control for variability in respondents’schedules and the possibility that time pressuremight affect work absorption. Earliest activity starttime was measured in elapsed minutes from mid-night (e.g., 1:00 AM equals 60 minutes).

Results

Given ourmultilevel data, we account for between-and within-person effects simultaneously in our pri-mary mediation analysis using multilevel structuralequation modeling (MSEM). The intraclass correla-tion coefficient for absorptionwas .46, indicating that46% of the variance was between person, whichsupports assessing both between- and within-personeffects. Table 2 displays the means, standard de-viations, and correlations for all variables.

Given Study 2’s primary aim to test the effects ofanticipated activities, we emphasize tests of Hy-potheses 2 through 5.4 Hypotheses 2 and 3 predicted

within-person effects of daily anticipated after-workactivities on daily work absorption. For this within-person analysis, there was no need to control forbetween-person variables (e.g., job type) since theydo not affect thewithin-person variance. However, itwas appropriate to control for day-level factors(e.g., earliest activity start time). Hypothesis 2 pre-dicted that anticipating domestic activities wouldhave a positive effect on work absorption. As showninTable 3, this hypothesis was supported (estimate50.50, SE5 0.22, p, .05). Hypothesis 3 predicted thatanticipating leisure activities would have a negativeeffect on work absorption. As shown in Table 3, thishypothesis was also supported (estimate 5 20.52,SE 5 0.23, p , .05).

Hypotheses 4 and 5 predicted indirect effects offamily structure on daily work absorption via theproportion of anticipated domestic (Hypothesis 4)and leisure (Hypothesis 5) activities. To appropri-ately test for indirect effects using data at variouslevels of analysis (between person5 Level 2; withinperson 5 Level 1), we tested the hypotheses withMSEM using Mplus, following the procedure out-lined by Preacher and colleagues for evaluatinga 2–1–1 mediation model (Preacher, 2011; Preacher,Zyphur, & Zhang, 2010). This approach decomposesthe variance into within- and between-person com-ponents, rather than consideringwithin- and between-person effects singularly (Preacher et al., 2010). In ourcase, the MSEM approach accounts for the effect of anindividual’s mean level of anticipated activities (be-tweenperson)ondailyabsorption (withinperson).Thebetween-level component reflects the relatively stableeffect of anticipated activities on absorption, so thatthe analysis captures the effect of an individual’stendency to anticipate a particular category of ac-tivities. The within-level component accounts fordaily variations in a person’s mean level of antici-pated activities on daily absorption. In this case, wecaptured day-to-day changes in anticipated activi-ties and the effects of variations from the mean.Given that family structure is a between-personvariable, both the between and within componentsprovide relevant information and are important ininterpreting theMSEMresults. Figure 1displays theMSEM model for this 2–1–1 analysis.

Hypothesis 4 predicted that anticipating domesticactivities after work would mediate the effect offamily structure on work absorption. As shown inTable 4, family structure had a significant negativeindirect effect on work absorption via anticipateddomestic activities (coefficient 5 2.17, SE 5 0.09,p# .05), thus supportingHypothesis 4. Hypothesis 5

4 A 2–1 multilevel analysis including between-personcontrol variables showed that single, childless people re-ported lower absorption (estimate52.41,SE5 0.22,p, .05,one tailed), consistent with Hypothesis 1. Although in-formative, this result should be interpreted with caution con-sidering the person-level sample size based on listwisedeletion (n5 892 8). Although ourN size is consistent withother diary studies testing multilevel effects (e.g., Ilies et al.,2011;Sonnentag&Zijlstra,2006), theperson-level samplesizeis relatively small for testing between-person effects only.


TABLE2

Intercorrelation

sbe

twee

nVariables

inStudy2

Personleve

lM

SD

23

45

67

89

1Singlech

ildless

0.26

0.44

2.03

2.19†

2.22*

2.02

2.28*

*.31*

*2.19†

2.10

2Gen

der

0.82

0.39

.10

2.09

2.01

.00

2.09

2.01

2.15

3Age

43.30

10.98

2.28*

*.18†

.17†

2.38*

*.20*

2.01

4Mon

thly

pay

1.88

0.32

.00

2.10

.13

2.17†

2.11

5Gen

eral

absorption

4.42

1.04

.04

2.06

.53*

*2.02

Dayleve

lM

SD

23

45

67

89

6Proportion

dom

estic

0.31

0.27

2.34*

*.35*

*2.05

7Proportion

leisure

0.14

0.11

2.23*

*2.31*

*.39*

*8

Daily

absorption

4.47

0.95

.01

2.05

.11

9Earliesta

ctivitystarttim

e10

02.24

132.11

.02

.05

.18*

*

Notes:C

orrelation

sat

theday

leve

lare

displaye

dbe

low

thediago

nal

(n5

439);correlation

sat

thepersonleve

l,av

erag

edac

ross

thefive

day

s,aredisplaye

dab

ovethediago

nal

(n5

97).nsizesareba

sedon

max

imum

like

lihoo

destimationmod

elsusingalla

vailab

ledata.

†p,

.10

*p

,.05

**p

,.01


predicted that anticipating leisure activities wouldmediate the effect of family structure on work ab-sorption. The indirect effect of family structure onabsorption via anticipated leisure activities was notsignificant (coefficient 5 2.17, SE 5 0.17, p . .05).Thus, Hypothesis 5 was not supported.

Testing for Alternative Explanations

It is possible that the overall quantity of antici-pated activities drives their indirect effect. In otherwords, our results could be interpreted to indicate

that employees with families are simply busier afterwork and must use their work time more efficiently,thus rendering them more focused at work. There-fore, testing for the effects of the sheer number ofactivities anticipated after work is important, andallows us to determine whether our findings aredriven simply by the quantity of anticipated activi-ties, rather than by their nature. Accordingly, wecounted the number of activities anticipated eachday and entered this variable as a mediator in thesame2–1–1mediationmodel used to test the effect ofanticipating domestic/leisure activities. The correlation

TABLE 3Effects of Proportion Domestic/Leisure Activities on Daily Work Absorption

Estimate SE Estimate SE

Intercept 4.72** 0.38 Intercept 5.04** 0.36Earliest activity 0.00 0.00 Earliest activity 0.00 0.00Proportion domestic 0.50* 0.22 Proportion leisure 20.52* 0.23

*p , .05**p , .01

FIGURE 1MSEM Model Showing a 2–1–1 Multilevel Mediation Model between Family Structure (FS), Planned

After-Work Activities (PA)—Domestic or Leisure, and Work Absorption (AB)

Within

PA AB

Observed FS PA AB

bw

cb

Between

ABFS

PA ab bb

Note: Figure is based on Preacher, Zhang & Zyphur (2011) and Hulsheger et al. (2013), p 316.


between the daily count of activities and daily ab-sorption was not significant (r 5 2.04, p . .05).Similarly, the indirect effect of number of activities(M 5 2.16, SD 5 0.84) was not significant (co-efficient 5 2.00, SE 5 0.03, p . .05). This findingsuggests that our effects are driven not by thenumber of anticipated after-work activities but bytheir nature (domestic or leisure).

Given the prominent place of positive affect intheorizing about the effects of family on work(Greenhaus & Powell, 2006; Rothbard, 2001), we alsoexamined whether the indirect effects remainedsignificant over and above any impact of positiveaffect from home spilling over into work. Therefore,we ran our 2–1–1 models controlling for positiveaffect in the personal/family domain. We used thePANAS measure of positive affect (Watson, Clark, &Tellegen, 1988). Participants read the instruction“Indicate the extent to which you have felt the fol-lowing at home or in reference to your personal lifein the past few weeks” followed by a list of ten ad-jectives (e.g., excited, proud) with ratings on a scalefrom 1, “very slightly or not at all,” to 5, “extremely,”which were averaged (M 5 3.6, SD 5 0.63, a 5 .86).The correlation between positive affect and dailyabsorption was not significant (r5 .16, p. .05). The

indirect effect of family structure on absorption wassignificant over and above positive affect when pro-portion of domestic activities was the mediator (co-efficient 5 2.17, SE 5 0.09, p # .05), but not whenproportion of leisure activities was the mediator(coefficient 5 2.20, SE 5 0.19, p . .05)—similar tothe results without the positive affect control. Wealso testedpositive affect as amediator of the effect offamily structure on absorption using a 2–2–1 modelin Mplus. The indirect effect was not significant(coefficient 5 2.01, SE 5 0.04, p . .05), suggestingthat our results are not explained by positive affect.

Another alternative explanation for our results isthat employees may be more absorbed at work whenfamily members rely on their incomes. To test this,we measured the extent to which respondents ful-filled a breadwinner role by averaging two itemstaken from Behson (2002): “I feel a great deal of re-sponsibility to provide for my household” and “If Ididn’t have my job, there would still be plenty ofmoney to support my household” (reverse coded)(rated on a scale from 1, “strongly disagree,” to 7,“strongly agree”; r5 .362, p, .001). The correlationbetween breadwinner status and daily absorptionwas not significant (r 5 2.004, p . .05). Controllingfor breadwinner status (M 5 5.76, SD 5 1.30), we

TABLE 4Multilevel Mediation Model (2-1-1) Predicting Daily Work Absorption from Family Structure and

Proportion Domestic/Leisure Activities

Mediator:Proportion Domestic Mediator:Proportion Leisure

Parameter Estimate SE Estimate SE

Between level

Intercept 2.13 1.58 2.51 1.71Path ab 20.16** 0.06 0.08* 0.04Path bb 1.02** 0.39 22.04 1.46Path cb 20.22 0.22 20.22 0.29Indirect effecta 20.17* 0.09 20.17 0.17Residual variance absorption 0.51** 0.11 0.52** 0.11Residual variance mediator 0.07** 0.01 0.01** 0.00

Within level

Path bw 0.01 0.19 20.22 0.19Residual variance absorption 1.00** 0.10 0.99** 0.10

Notes: Models are fixed slope models; n 5 97 at the person level and n 5 439 at the day level, based on maximum likelihood estimationmodels using all available data. Average number of observations per person5 4.56. The paths refer to those shown in Figure 1. Path ab, familystructure → proportion domestic; path bb, proportion domestic → absorption; path cb, family structure → absorption; path bw, proportiondomestic/leisure → absorption. The model includes the following control variables: sex, age, monthly pay, general absorption, and earliestactivity start time.

a The indirect effect is represented in Figure 1 as the paths ab and bb.*p , .05

**p , .01


found that the indirect effect of family structure onabsorption was significant when the mediator wasproportion of domestic activities (coefficient52.17,SE5 0.09,p# .05), but notwhen itwas proportion ofleisure activities (coefficient52.17, SE5 0.17, p..05). Consistent with our primary analysis, these re-sults support our hypothesis that anticipated activi-ties explain the effects of family structure on workabsorption over and above breadwinner status.We also tested breadwinner status as a mediator ofthe relationship between family structure and ab-sorption. The indirect effect was not significant (co-efficient 5 2.01, SE 5 0.04, p . .05), furtherdiscounting this alternative explanation of our results.

Discussion

The Study 2 data allowed us to test our predictionthat anticipating after-work activities explains theeffect of family structure on work absorption. Aspredicted, we found that anticipatingmore domesticactivities led to increased work absorption, whereasanticipatingmore leisure activities led to lowerworkabsorption. Also as predicted, we found a significantindirect effect of family structure onwork absorptionoperating through the daily anticipation of domesticactivities afterwork, but not through the anticipationof leisure activities.

Study 2 shed additional light on the relationshipbetween family structure, nonwork activities, andwork absorption. Moreover, the daily diary surveydesign of Study 2 strengthens the causal inference ofour effects suggesting that, in fact, anticipating cer-tain nonwork activities does predict the individual’sabsorption at work. Additionally, our alternativeanalyses suggest that the nature of participants’ after-work activities may affect work absorption over andabove the sheer number of nonwork activitiesplanned, positive affect, or breadwinner status—all plausible alternative explanations for our effects.Also, contrary to expectations based on the idealworker norm that those without families are “un-encumbered,” there was no significant differencebetween single, childless employees and others inthe overall number of anticipated after-workactivities.

STUDY 2: SUPPLEMENT

The primary aim of the supplement to Study 2was to investigate whether respondents construedtheir after-work activities consistently with our the-orizing. We also sought to provide supplemental

evidence that their construal of after-work activitiesmediates the effect of anticipated activities on ab-sorption. Therefore, we administered surveys to asample of full-time employees regarding their after-work activities, their construal of those activities,and their work absorption.

Sample and Procedures

Participants were Qualtrics panel members—working adults who agree to complete onlinesurveys to earn points they can redeem for mer-chandise online. Only full-time employees whocorrectly responded to attention filter items(e.g., “Leave this question blank”)were included inthe sample. Similarly to Study 2, on a morningsurvey, respondents reported their anticipatedafter-work activities. Respondents also reportedtheir construal of each activity in the morning.Given our theorizing that anticipation of after-work activities impacts absorption because of theway people construe those activities, we thought itmost appropriate to measure activity construal atthe time of anticipation. On an afternoon survey,respondents reported their work absorption.5

Three hundred and eighty participants completedthe morning survey, but 123 responded that theyhad no plans after work, so they were excludedsince the analysis focused on construal of antici-pated activities. Our final sample included 196participants with matched morning and afternoonsurveys, reporting a total of 564 activities, of which429 were categorized as domestic or leisure.

Measures

Ourmeasure of absorption (a5 .90) and categoriesof activities were identical to those used in Study 2.Participants also rated each activity on the extent towhich it was goal directed, obligatory, or an escapefrom the routine. Goal direction was measured withfour items (a5 .92), modified fromWoolley’s (2009)measure of outcome focus. A sample item was“When participating in this activity, I think about

5 We collected data a second day to compare the effectsof recovery due to the prior day’s activities to the effects ofsame-day anticipated activities. This supplemental analysisis speculative, given that we collected recovery measuresand anticipated activities on the same survey, potentiallyshaping the results. However, we address these results inthe General Discussion section, and details of this analysisare available from the authors upon request.


what the final outcome of this activity will be” (ratedfrom 1, “strongly disagree,” to 7, “strongly agree”).To measure the extent to which activities wereconstrued as obligatory, participants answeredfive questions (a 5 .81), adapted from Weinsteinand Ryan’s (2010) controlled motivation scale. Asample item was “I participate in this activity be-cause I feel I have to” (rated on a scale from 1, “notat all true,” to 7, “very true”). To measure escapefrom the routine, participants answered threequestions (a 5 .93), modified from the escapingpressures subscale of Schulz and Watkins’s (2007)leisure meanings inventory. A sample item was“Forme, this activity is a break, a change from life’susual routine” (rated from 1, “strongly disagree,”to 7, “strongly agree”).

Results

To assess differences in how respondents con-strued domestic and leisure activities at the activitylevel, we used the mixedmodel function in SPSS, asthe after-work activities (Level 1) are nested withinpeople (Level 2). To test for the effects of activityanticipation on absorption mediated by construal,weused thePROCESSmacro for SPSS (Hayes, 2013).Table 5 displays the means, standard deviations,and correlations for all variables.

Construal of leisure and domestic activities.Ourtheorizing assumes that anticipated domestic activ-ities are construed as more obligatory, more goal di-rected, and less of an escape from the routine relativeto leisure activities, which are voluntary (i.e., lessobligatory), pursued for intrinsic reasons (i.e., lessgoal directed), and an escape from the daily routine.There was a significant effect of category type

(domestic, leisure, or other) on construal of theextent to which activities were goal directed,F(2, 537.60)522.83,p, .001), obligatory,F(2, 511.02)5 48.56, p , .001, and an escape from the routine,F(2, 562.04)5 115.70,p, .001.As expected, pairwisecomparisons revealed that domestic activities wereconstrued as more goal directed (M 5 4.01) thanleisure activities (M 5 3.42, p , .001), and moreobligatory (M 5 3.80) than leisure activities (M 52.57, p , .001). Also as expected, leisure activitieswere perceived as more of an escape from theroutine (M 5 5.75) than were domestic activities(M 5 3.27, p , .001).

Construal as mediator. Next, we explored theextent to which activity construal mediated the ef-fects of anticipating activities on afternoon absorp-tion. Respondents indicated how long they expectedto participate in each after-work activity, and wecalculated the proportion of anticipated domesticand leisure time planned for after work, an im-provement over the count-based proportions used inStudy 2. We aggregated each construal to the personlevel by averaging them for all activities anticipatedthat day. For example, if a respondent anticipatedthree activities, we averaged their ratings of goaldirection across activities to reflect their overallconstrual of goal direction for theday.AswithStudy2, we controlled for earliest activity start time, joblevel, age, and gender (collected one day before thesurveys about activities).

Our theorizing suggests that, because domesticactivities are construed as obligatory, goal directed,and routine, anticipating them reinforces workabsorption. Indeed, the indirect effect of anticipat-ing domestic activities on absorption through themeasure of obligatory construal is positive and

TABLE 5Means, Standard Deviations, and Correlations for Study 2 Supplement

Variable Mean SD 1 2 3 4 5 6 7 8 9

1 Absorption 4.08 1.482 Proportion leisure 0.31 0.40 2.053 Proportion domestic 0.38 0.40 .00 2.51**4 Obligatory 3.22 1.33 .13† 2.27** .29**5 Escape from routine 4.41 1.67 .19** .43** 2.41** 2.106 Goal directed 3.90 1.42 .23** 2.27** .08 .41** .117 Age 45.44 12.25 .07 2.03 .06 2.13† 2.01 2.058 Gender 0.72 0.45 2.11 2.04 .07 .02 2.09 .03 2.119 Job level 2.94 0.76 .11 2.11 .09 .01 .06 2.01 .29** 2.20**10 Earliest activity start time 1667.96 270.37 .03 2.01 2.13† .01 .09 .00 2.05 .00 .06

n 5 196.†p , .10

**p , .01


significant (estimate5 0.18; 95% CI [0.02, 0.42]; seeTable 6). With goal direction as the mediator, theindirect effect was not significant (estimate 5 0.08;95% CI [20.02, 0.24]; see Table 6). With escapefrom the routine as the mediator, the indirect effectwas significant but negative (estimate520.32; 95%CI [20.67, 20.07]; see Table 6), contrary to ourexpectations.

Because leisure activities are construed as volun-tary (rather than obligatory), intrinsically motivated(rather than goal directed), and an escape from theroutine, we expected that anticipating leisure activ-ities would reduce work absorption. As expected,with obligatory construal as the mediator, the in-direct effect of anticipating leisure activities on ab-sorption was negative and significant (estimate 520.14; 95% CI [20.37, 20.01]; see Table 6), as wasthe indirect effect with goal direction as themediator(estimate 5 20.24; 95% CI [20.52, 20.08]; seeTable 6). Contrary to our expectations, however,with escape from the routine as the mediator, theindirect effect was positive and significant (estimate50.40; 95% CI [0.14, 0.75]).

Testing for Alternative Explanations

As with Study 2, we sought to understand the roleof positive affect in our findings, given the impor-tance of affect to research on work engagement andthe work–nonwork interface (e.g., Greenhaus &Powell, 2006; Rothbard, 2001). Therefore, we col-lected respondents’ positive affect on the same sur-vey with which we collected data on anticipatedactivities, using the same items from Study 2, mod-ified to address affect on the day of interest. Specif-ically, the item stem stated “Indicate the extent towhich you feel the following today,” followed byaffect descriptors (e.g., excited, distressed) (a5 .92).Our analyses of the supplemental data show thatneither anticipating domestic nor leisure activities

was significantly related to positive affect (domestic:r 5 2.12, p . .05; leisure: r 5 .02, p . .05). Further,there was no significant indirect effect of domesticactivities on absorption through positive affect (es-timate520.20; 95%CI [20.45, 0.01]), nor was therea significant indirect effect of leisure activities onabsorption through positive affect (estimate5 0.048;95% CI [20.16, 0.27]). These analyses suggest thatpositive affect does not drive our results. Further,these findings reinforce the perspective that non-work experiences can help increase work absorptionregardless of whether the experiences are explicitlypositive.

Discussion

Theanalysis of activity construal provides supportfor key assumptions in our theorizing. That is, do-mestic activities are seen as more obligatory, moregoal directed, and less of an escape from the routinethan leisure activities. Further, as we expected,construing activities as more goal directed andobligatorywas positively related towork absorption.The analyses of indirect effects suggest that con-struing activities as obligatory drives the effects onabsorption for both domestic and leisure activities.Goal direction, however, appears to be relevant onlyfor explaining why leisure activities (which are lowin goal direction) have the predicted negative effecton absorption.

Although the logic of anticipatory transitions(Richter, 1990) and current concerns theory (Klingeret al., 1980) suggest that anticipating an escape fromthe daily work routine should detract from workabsorption, the positive effect of construing an ac-tivity as an escape from the routine suggests someadditional complexities that may not be capturedhere. We offer several possibilities for future re-search to explore. In particular, characterizing anactivity as “an escape from the daily routine” does

TABLE 6Indirect Effects of Anticipated Activities on Absorption: Study 2 Supplement

Independent Variable:Proportion Domestic Independent Variable:Proportion Leisure

Mediator Estimate LLCI ULCI Mediator Estimate LLCI ULCI

Obligatory 0.18(.09) 0.02 0.42 Obligatory 20.14(.09) 20.37 20.01Goal direction 0.08(.08) 20.02 0.24 Goal direction 20.24(.11) 20.52 20.08Escape 20.32(.15) 20.67 20.07 Escape 0.40(.15) 0.14 0.75

Notes:Allmodels include the followingcontrol variables: age, gender, job level, andearliest activity start time. Each indirect effectwas runasa separate model. Estimates reflect activity→mediator→ absorption effects. Values in bold type show significant indirect effects (CI excludeszero). LLCI 5 lower level of 95% bootstrapped confidence interval; ULCI5 upper level of 95% bootstrapped confidence interval.


not account for the fact that anticipating some non-routine activities (e.g., watching a television special)may not be sufficiently compelling to distract fromthe current task and reduce absorption—particularlyfor someone performing complex, engaging work.This idea is consistent with recovery researchshowing that features such as how active, passive, orchallenging an activity is (Fritz & Sonnentag, 2006;Rook & Zijlstra, 2006) may shape the extent to whichemployees detach and recover fromwork.Moreover,there may be an interplay between an activity’snonroutine nature and other characteristics, such asgoal direction. For instance, playing in a competitivesoccer game after workmay be a break from the dailyroutine, but it may also be construed as a goal-directed activity—particularly for someone moti-vated to win or reach a certain performance level inthe game—thus potentially reinforcing a workmindset.

In sum, although our unexpected finding suggeststhat anticipating an escape from the routine mayprovide a boost in resources available for work ab-sorption in the same way that taking a break afterwork provides recovery, perhaps only certain non-routine activities—for example, those that are highlyactive and low in goal direction—cause the type ofdistraction we predicted. Importantly, our resultswere almost entirely as predicted: anticipating lei-sure activities had a negative effect on work absorp-tion through their negative relationship with goaldirection and sense of obligation. In total, those ef-fects may supersede the positive effects of viewingleisure activities as an escape from the daily routine.Moreover, our overall findings—that anticipatingafter-work domestic activities explains the effect offamily structure on absorption—support our argu-ment that family responsibilities can lead to in-creased absorption at work.

GENERAL DISCUSSION

We examined the relationship between em-ployees’ family structure and work absorption,and how anticipated after-work activities mediatethis relationship. Supporting our hypothesis, wefound in our first study that single, childless em-ployees reported lower work absorption than em-ployees with other family structures combined.Our second study revealed a negative indirect ef-fect of being single and childless on work absorp-tion via employees’ anticipated after-workactivities. As we predicted, anticipating domesticactivities after work increased work absorption,

whereas anticipating leisure activities decreasedwork absorption. Further, we found that anticipat-ing domestic activities after work mediated the ef-fect of family structure on absorption, whereasanticipating leisure activities did not. Finally, ourfindings suggest that the differential construal ofdomestic and leisure activities as goal directed andobligatory helps to explain the effects of anticipatedactivities on absorption. In sum, our findings sug-gest that the daily tasks associated with havinga spouse and/or children induce greater psycho-logical immersion into the work role.

Expanding Theories for How the Nonwork DomainAffects Work

Our findings contribute to existing theoreticalexplanations by showing that the presence of a nu-clear family can induce greater focus at work. His-torical assumptions that family responsibilitiesdetract from work remain prevalent (Kelly et al.,2010; Reid, 2015; Williams, 2001). For example, ina recent survey of more than 2,000 employees, 40%agreed thatworkerswithout personal commitmentsare most productive, and the majority believedthat ideal workers would make themselves avail-able for business needs regardless of the time of day(SHRM Online Staff, 2011). Some research doesshow that having a family can result in increaseddemands at home, and that, for some employees,family responsibilities can negatively impact work(e.g., Frone et al., 1992; Rothbard, 2001)—findingsthat may help to fuel negative work–family as-sumptions. In contrast, and consistent with the en-richment perspective, we consider how family mayaffect work positively. Unlike those of prior en-richment studies, our findings suggest that partici-pation in the family role canyield positive effects onwork whether or not the connection between workand family is characterized by positive affect or skilltransfer between the two roles—fundamental tenetsof enrichment theorizing (Greenhaus & Powell,2006; Rothbard, 2001).

Our work also provides a different perspective onresearch showing that participation in leisure activ-ities after work enhances recovery from work, thusenabling heightenedwork engagement the followingday (e.g., Sonnentag & Zijlstra, 2006). Ironically, wefound that anticipating the activities less likely tofacilitate recovery after work (domestic) keeps em-ployees more psychologically immersed in theirwork during the day. Conversely, anticipating ac-tivities more likely to provide recovery after work


(leisure) induces psychological detachment fromwork before the workday is done, resulting in lowerwork absorption. Importantly, in an alternativeanalysis of the Study 2 supplement data, we foundthat anticipating leisure activities significantly re-duced afternoon absorption even when taking intoaccount the prior day’s activities and morning re-covery. This finding raises interesting questionsabout the net effects of after-work leisure activitieson employees, and it suggests that the period of timebefore a leisure activity may be just as important toconsider in recovery research as the period of timeafter the activity. Our work also illuminates a re-lationship between family structure and after-workactivities that may help scholars understand how anemployee’s family structure may constrain or en-hance their ability to recover fromwork at the end ofthe day. By connecting absorption, anticipation ofafter-work activities, and family structure, we in-troduce factors that can potentially expand the logicof recovery research.

Finally, our studies contribute to both theory andpractice by focusing on single, childless workers,a growing portion of the population largely un-addressed in work–nonwork theorizing (Eby et al.,2005). In our data, the lower reported work absorp-tion of single, childless employees was explained byfewer domestic activities scheduled after work,a finding consistentwith expectations that their livesmay be less encumbered than those of other em-ployees.However, although theyanticipated a greaterproportion of leisure activities than workers withother family structures, this did notmediate the effectof family structure on absorption. Further, single,childless workers reported no fewer activities overallthan those with families, and the total number of ac-tivities planned for after work did not mediate therelationship between family structure and work ab-sorption. These findings highlight the need to con-sider more closely the effects of non-family nonworkactivities on employees’work experiences. They alsoilluminate our understanding of the nonwork activi-ties of single, childless workers.

Single childless workers exist at the intersectionof ideal worker expectations that potentially rendertheir family structure a positive signal to organiza-tions and pressure to live up to the “ideal worker”norm, as our opening quote illustrates. Hamilton andcolleagues (2006) found that single, childless womenexperiencedwork–life conflict due to these pressuresand expectations that they would be continuouslyavailable for work. Additionally, many single work-ers often feel that they do not have organizational

support in pursuing fulfilling nonwork lives (Casper,Weltman,&Kwesiga, 2007) and feel overlookedwhencompanies only offer family-specific policies (Grover,1991; Kossek & Nichol, 1992). Thus, although single,childless workers may benefit from assumptions ofbeing unencumbered and constantly accessible,many experience these expectations as unwantedpressure or an undue burden (Casper et al., 2007;Rothausen, Gonzalez, Clarke, & O’Dell, 1998). Single,childless workers in our sample reported more de-velopmental and community volunteering activitiesthan did workers with other family structures—further reinforcing the idea that they seek full andactive lives outside of work rather than singular de-votion and constant accessibility to the work organi-zation. Moreover, their work–life interface, needs,and concerns are likely more complex than is com-monly acknowledged, thus warranting further study.

Practical Implications for Managersand Employees

Our focus on the psychological effects of antici-pating future activities presents both a new way ofthinking about the connection between work andnonwork and also suggests action steps that bothindividuals and organizations can take to leverageemployees’ nonwork responsibilities. In our data,few respondents anticipated only leisure activitiesand few anticipated only domestic activities, so wesuggest that selectively invoking specific after-workactivities may be a way to enhance absorption.Boundary theorists document instances in whichthoughts of nonwork roles during the workday candistract workers and disrupt their work mindset(e.g., Glavin et al., 2011; Smit et al., 2016). However,depending on how employees manage the work–nonwork boundary, these cognitive transitions tononwork roles can also induce a greater focus onwork (Richter, 1990). Nippert-Eng (1996: 572), in herseminal work on work–nonwork boundary man-agement, described the ways employees could usethe placement of calendars, to-do lists, or visual re-minders of upcoming tasks to redirect their focus:

This calendar lists three types of events: her vacationdays, her departmental softball games, and weekendvisits fromher family. . . . Joan simply has to lookup tobe reminded that, at a certain time, she’ll be doingsomething she really enjoys, something more per-sonally exciting, perhaps, than the task at hand.

According to our findings, this display of leisureactivities likely reduces the employee’s work


absorption, but a display of future domestic tasksmight reinforce a work focus. Overall, Nippert-Eng’s(1996) work suggests that the choice to display re-minders of future events selectively and strategicallymay help employees with both planning and refo-cusing on work. This is consistent with Richter’s(1990) concept of anticipatory boundary transitions,and her theorizing that anticipating domestic activi-ties after work reinforces a work mindset. Addition-ally, Mainemelis (2001) suggested that organizationalrituals (e.g., morning coffee in the break room) canhelp employees transition to and from work in waysthat contribute to work absorption. Moreover, mind-fulness training can help employees minimize un-planned mental transitions to nonwork roles andfocus more at work (Michel, Bosch, & Rexroth, 2014).

Our results also suggest that managers may need toattend more closely to the work–life needs of single,childless employees. Casper et al. (2007) found thatsingle, childless workers felt more supported in orga-nizations that provided them equal benefits andshowed equal respect for diverse aspects of em-ployees’ nonwork lives. Organizations should ac-knowledge the complexity of the work lives of single,childless workers—perhaps adopting policies thathelp them pursue goal-directed activities outside ofwork. Such organizational support could help makeworkmorecompelling for single, childless employees.Importantly, our results do not suggest that organiza-tions benefit from having employees who are un-satisfiedorunstimulated in their lives outside ofwork.Rather, we recommend that organizations work tounderstand the diversity of employees’ nonwork livesand create environments that value those differences.

Limitations and Future Directions

Although our research contributes to the literaturein important ways, we must acknowledge its limi-tations. One limitation is that the survey data con-sisted of self-reportedbehaviors, potentially yieldingcommon method variance (Podsakoff et al., 2012).Importantly, however, the respondents themselvesare the best source of information for our key con-structs. Moreover, given that our independent variable—family structure—is composed of respondents’ maritaland parental status rather than attitudes or perceptions,collecting it on the survey with work absorption is notlikely to inflate statistical effects between the twovariables. Finally, we collected the Study 2 data atmultiple points in time, and the results were consis-tent with the Study 1 results, further reducing con-cerns over common method bias.

Important questions unanswered by our analysesconcern the effects of spousal employment status andthe employee’s gender, two critical factors in the tra-ditional “ideal worker” ideology, which rests on themodel of a male employee with a stay-at-home wife(Williams, 2001). Moreover, our Study 2 and supple-ment samples are predominantly female, potentiallyshaping our results, although Study 1 was more gen-der balanced. To address these issues, we conductedadditional analyses decomposing family structureinto six mutually exclusive categories, crossing mar-ital status, employment status of the spouse, and pa-rental status. These analyses allowed us to comparethe single, childless category to each of the otherfamily categories (e.g., married/unemployed spouse/no child, married/employed spouse/child, etc.). Theresults of this post-hoc analysis of covariance, al-though only suggestive due to inconsistent cell sizesacross categories, provide preliminary evidence thatspousal employment status is not a systematic driverof our results. Importantly, this analysis also indicatesthat our effects are not driven simply by parental ormarital status alone. In addition, our Study 2 findingssuggest that domestic tasks affect work absorptionover and above gender.6 Futurework should continueto explore the effects of different family structures,including cohabitating heterosexual couples (notstudied here), as well as gender differences.

Another important issue to explore in future re-search is the overall impact of after-work activitieson work. Whereas our findings present an interest-ing contrast to research on after-work recovery(Sonnentag & Zijlstra, 2006), they also raise thequestion of how the effects of anticipating after-workactivities relate to theafter-effectsof thoseactivitieson the following day. Yet, they also show that leisureactivities are construed as an escape from the dailyroutine, and anticipating this escape positively af-fects absorption, a finding consistent with a funda-mental tenet of recovery research—that detachingfrom work is beneficial. These findings, consideredalongwith our overall negative effects of anticipatingleisure activities on work absorption, suggest thatthere are remaining complexities to explore, and thatmoderators—such as how active (Rook & Zijlstra,2006) or challenging (Fritz & Sonnentag, 2006) theleisure activity is—should be examined. A review of

6 In an alternative analysis not shown, we found thatthe interaction between family structure and sex (whichserved as a proxy for gender) did not have a significanteffect on absorption (p . .05), suggesting that the maineffect for family structure did not depend on gender (sex).


the activities anticipated by our respondents (seeAppendix A) suggests that many of them were nothighly challenging or active, but we do not have dataon respondents’ perceptions of their activities onthese particular dimensions. However, those partic-ular aspects of after-work activities could be an im-portant issue for scholars to explore. Overall, futureresearch should continue to consider the effects ofafter-work activities comprehensively—how theyaffect the work experience both before and afterwork, which effect is strongest, and how differentfeatures of after-work activities shape these effects.

Finally, we examined respondents’ work absorp-tion rather than time spent at work or productivity.Our examination of absorption in the context of theideal worker ideology emphasizes the view that anemployee’s value in the organization may not be pri-marily dependent upon constant accessibility andface-time spent atwork, but, rather, on an individual’sability to give his or her maximum mental energywhile there. This is particularly the case for pro-fessional knowledge workers. Many organizationsbuy into the “ideal worker” ideology when they re-ward “face time” either implicitly or explicitly, andthey send the message to employees that they aremore valuable to the extent that they can makethemselves available to work on an unlimited basis(Reid, 2015). This benefits employees with an “idealworker” family structure, but there is little evidencethat those who spend more time at work are morevaluable. Future research considering how familystructure affects a variety of other work outcomeswould help expand theory regarding the effect ofnonwork on work.

CONCLUSION

The evolution of family structure has caused ashift in organizational demographics, prompting areexamination of our expectations about the work–family relationship. A critical part of revising currenttheory about the effect of family onwork is to considerfamily structure more comprehensively, and also toexamine directly the way family structure shapes peo-ple’s nonwork lives.Overall,with thiswork,we seek toenrich our collective knowledge of the effects of familyresponsibilities on employees’work experiences.

REFERENCES

Aryee, S., & Luk, V. 1996. Balancing two major parts ofadult life experience:Work and family identity amongdual-earner couples. Human Relations, 49: 465–487.

Ashforth, B. E., Kreiner, G. E., & Fugate, M. 2000. All ina day’s work: Boundaries and micro role transitions.Academy of Management Review, 25: 472–491.

Ashforth, B. E., &Mael, F. 1989. Social identity theory andthe organization. Academy of Management Review,14: 20–39.

Avery, D. R., McKay, P. F., & Wilson, D. C. 2007. Engagingthe aging workforce: The relationship between per-ceived age similarity, satisfaction with coworkers,and employee engagement. The Journal of AppliedPsychology, 92: 1542–1556.

Bailyn, L., Drago, R.W., & Kochan, T. A. 2001. Integratingwork and family life: A holistic approach. A reportof the Sloan Work–Family Policy Network. Boston,MA: Sloan School of Management, Massachusetts In-stitute of Technology. Retrieved from http://web.mit.edu/workplacecenter/docs/WorkFamily.pdf

Behson, S. J. 2002. Copingwith family-to-work conflict: Therole of informal work accommodations to family. Jour-nal of Occupational Health Psychology, 7: 324–341.

Bittman, M., & Wajcman, J. 2000. The rush hour: Thecharacter of leisure time and gender equity. SocialForces, 79: 165–189.

Blair-Loy, M. 2004. Work devotion and work time. InC. F. Epstein & A. L. Kalleberg (Eds.). Fighting fortime: Shifting boundaries of work and social life:216–282. New York, NY: SAGE.

Casper, W. J., Weltman, D., & Kwesiga, E. 2007. Beyondfamily-friendly: The construct and measurement ofsingles-friendly work culture. Journal of VocationalBehavior, 70: 478–501.

Collins, G. 2008, December 4. One singular sensation. TheNew York Times, Available at: https://www.nytimes.com/2008/12/04/opinion/04collins.html.

Coverman, S. 1989. Role overload, role conflict, and stress:Addressing consequences of multiple role demands.Social Forces, 67: 965–982.

Csikszentmihalyi, M., & LeFevre, J. 1989. Optimal experi-ence in work and leisure. Journal of Personality andSocial Psychology, 56: 815–822.

Dillman, D. A. 2000. Mail and internet surveys: The tai-lored design method. New York, NY: Wiley.

Dumas, T. L., & Sanchez-Burks, J. 2015. The professional,the personal, and the ideal worker: Pressures and ob-jectives shaping the boundary between life domains.The Academy of Management Annals, 9: 803–843.

Dumas, T. L., & Stanko, T. L. 2017. Married with children:How family role identification shapes leadership be-haviors at work. Personnel Psychology, 70: 515–633.

Eby, L. T., Casper, W. J., Lockwood, A., Bordeaux, C.,& Brinley, A. 2005. Work and family research inIO/OB: Content analysis and review of the literature


http://web.mit.edu/workplacecenter/docs/WorkFamily.pdf

http://web.mit.edu/workplacecenter/docs/WorkFamily.pdf

https://www.nytimes.com/2008/12/04/opinion/04collins.html

https://www.nytimes.com/2008/12/04/opinion/04collins.html

(1980–2002). Journal of Vocational Behavior, 66:124–197.

Edwards, J. R., &Rothbard,N. P. 2000.Mechanisms linkingwork and family: Clarifying the relationship betweenwork and family constructs. Academy of Manage-ment Review, 25: 178–199.

Etzion, D., Eden, D., & Lapidot, Y. 1998. Relief from jobstressors and burnout: Reserve service as a respite.The Journal of Applied Psychology, 83: 577–585.

Forster, S., & Lavie, N. 2014. Distracted by your mind?Individual differences in distractibility predict mindwandering. Journal of Experimental Psychology.Learning, Memory, and Cognition, 40: 251–260.

Fritz, C., & Sonnentag, S. 2006. Recovery, well-being, andperformance-related outcomes: The role of workloadand vacation experiences. The Journal of AppliedPsychology, 91: 936–945.

Frone, M. R., Russell, M., & Cooper, M. L. 1992. Anteced-ents and outcomes of work–family conflict: Testinga model of the work–family interface. The Journal ofApplied Psychology, 77: 65–78.

Glavin, P., Schieman, S., & Reid, S. 2011. Boundary-spanning work demands and their consequences forguilt and psychological distress. Journal of Healthand Social Behavior, 52: 43–57.

Gordon, J. R., &Whelan, K. S. 1998. Successful professionalwomen in midlife: How organizations can more effec-tively understand and respond to the challenges. TheAcademy of Management Executive, 12: 8–24.

Graves, L. M., Ohlott, P. J., & Ruderman, M. N. 2007.Commitment to family roles: Effects on managers’attitudes and performance. The Journal of AppliedPsychology, 92: 44–56.

Greenhaus, J. H., & Beutell, N. J. 1985. Sources of conflictbetween work and family roles. Academy of Man-agement Review, 10: 76–88.

Greenhaus, J. H., & Powell, G. N. 2006. When work andfamily are allies: A theory ofwork–family enrichment.Academy of Management Review, 31: 72–92.

Grover, S. l. 1991. Predicting the perceived fairness ofparental leave policies. The Journal of Applied Psy-chology, 76: 247–255.

Haas, M. R., Criscuolo, P., & George, G. 2015. Whichproblems to solve? Online knowledge sharing and at-tention allocation in organizations. Academy of Man-agement Journal, 58: 680–711.

Hamilton, E. A., Gordon, J. R., &Whelan-Berry, K. S. 2006.Understanding the work–life conflict of never-married women without children. Women in Man-agement Review, 21: 393–415.

Hammer, L. B., Neal, M. B., Newsom, J. T., Brockwood,K. J., & Colton, C. L. 2005. A longitudinal study of the

effects of dual-earner couples’ utilization of family-friendly workplace supports on work and familyoutcomes. The Journal of Applied Psychology, 90:799–810.

Hayes, A. F. 2013. Introduction tomediation,moderation,and conditional process analysis: A regression-based approach. New York, NY: The Guilford Press.

Hobfoll, S. E. 2011. Conservation of resources theory: Itsimplication for stress, health, and resilience. InS. Folkman (Ed), The Oxford handbook of stress,health, and coping: 127–147. New York, NY: OxfordUniversity Press.

Hochschild, A. R. 1997. The time bind: When work be-comes home and home becomes work. New York,NY: Metropolitan Books.

Hoobler, J. M., Wayne, S. J., & Lemmon, G. 2009. Bosses’perceptions of family–work conflict and women’spromotability: Glass ceiling effects. Academy of Man-agement Journal, 52: 939–957.

Hulsheger, U. R., Alberts, H. J., Feinholdt, A., & Lang,J.W. 2013. Benefits of mindfulness at work: the roleof mindfulness in emotion regulation, emotionalexhaustion, and job satisfaction. Journal of AppliedPsychology, 98: 310–325.

Ilies, R., Keeney, J., & Scott, B. A. 2011. Work–family in-terpersonal capitalization: Sharing positive workevents at home. Organizational Behavior and Hu-man Decision Processes, 114: 115–126.

Iso-Ahola, S. 1997. A psychological analysis of leisureand health. In J. T. Haworth (Ed.), Work, leisure andwell-being: 131–144. London, U.K.: Routledge.

Kahn, W. A. 1990. Psychological conditions of personalengagement and disengagement at work. Academyof Management Journal, 33: 692–724.

Kelly, E. L., Ammons, S.K., Chermack,K., &Moen, P. 2010.Gendered challenge, gendered response confrontingthe ideal worker norm in a white-collar organization.Gender & Society, 24: 281–303.

Klinger, E., Barta, S. G., & Maxeiner, M. E. 1980. Motiva-tional correlates of thought content frequency andcommitment. Journal of Personality and SocialPsychology, 39: 1222–1237.

Kossek, E. E., &Nichol, V. 1992. The effects of on-site childcare on employee attitudes and performance. Per-sonnel Psychology, 45: 485–509.

Kroshus, E. 2008. Gender, marital status, and commer-cially prepared food expenditure. Journal of Nutri-tion Education and Behavior, 40: 355–360.

Lee, Y. G., & Bhargava, V. 2004. Leisure time: Do marriedand single individuals spend it differently? Familyand Consumer Sciences Research Journal, 32:254–274.


Lee, B., Lawson, K. M., Chang, Po-Ju, Neuendorf, C., &Dmitrieva, N. O. 2015. Leisure-time physical activitymoderates the longitudinal associations betweenwork–family spillover and physical health. Journalof Leisure Research, 47: 444–466.

Mainemelis, C. 2001. When the muse takes it all: A modelfor the experience of timelessness in organizations.Academy of Management Review, 26: 548–565.

Martins, L. L., Eddleston, K. A., & Veiga, J. F. 2002. Moder-ators of the relationship between work–family conflictand career satisfaction. Academy of ManagementJournal, 45: 399–409.

Meijman,T. F., &Mulder,G. 1998. Psychological aspects ofworkload. In P. J. D. Drenth, H. Thierry & C. J. deWolff(Eds.),Handbook of work and organizational, vol. 2:Work psychology (2nd ed.): 5–33. Hove, U.K.: Psy-chology Press/Erlbaum (U.K.) Taylor & Francis.

Menges, J. I., Tussing, D. V., Wihler, A., & Grant, A. M.2017. When job performance is all relative: Howfamily motivation energizes effort and compensatesfor intrinsic motivation. Academy of ManagementJournal, 60: 685–719.

Michel, A., Bosch, C., & Rexroth, M. 2014. Mindfulness asa cognitive–emotional segmentation strategy: An in-tervention promoting work–life balance. Journal ofOccupational and Organizational Psychology, 87:733–754.

Nippert-Eng, C. E. 1996. Home and work: Negotiatingboundaries through everyday life. Chicago, IL: Uni-versity of Chicago Press.

Nomaguchi, K. M., & Bianchi, S. M. 2004. Exercise time:Gender differences in the effects of marriage, parent-hood, and employment. Journal of Marriage and theFamily, 66: 413–430.

Oerlemans, W. G. M., Bakker, A. B., & Demerouti, E. 2014.How feeling happy during off-job activities helpssuccessful recovery from work: A day reconstructionstudy. Work and Stress, 28: 198–216.

Podsakoff, P.M.,MacKenzie, S. B., & Podsakoff, N. P. 2012.Sources of method bias in social science research andrecommendations on how to control it. Annual Re-view of Psychology, 63: 539–569.

Preacher, K. J. 2011. Multilevel SEM strategies for evalu-ating mediation in three-level data. Multivariate Be-havioral Research, 46: 691–731.

Preacher, K. J., Zyphur, M. J., & Zhang, Z. 2010. A gen-eral multilevel SEM framework for assessing mul-tilevel mediation. Psychological Methods, 15:209–233.

Preacher, K. J., Zhang, Z., & Zyphur,M. J. 2011. Alternativemethods for assessing mediation in multilevel data:The advantage of multilevel SEM. Structural Equa-tion Modeling, 18: 161–182.

Quinn, R. W. 2005. Flow in knowledge work: High per-formance experience in the design of national securitytechnology. Administrative Science Quarterly, 50:610–641.

Reid, E. 2015. Embracing, passing, revealing, and the idealworker image: How people navigate expected andexperienced professional identities. Organization Sci-ence, 26: 997–1017.

Richter, J. 1990. Crossing boundaries between professionalandprivate life. InH.Y.Grossman&N.L.Chester (Eds.),The experience and meaning of work in women’slives: 143–163. Hillsdale, NJ: L. Erlbaum Associates.

Rodriguez-Sanchez, A. M., Schaufeli, W., Salanova, M.,Cifre, E., & Sonnenschein, M. 2011. Enjoyment andabsorption: An electronic diary study on daily flowpatterns. Work and Stress, 25: 75–92.

Rook, J. W., & Zijlstra, F. R. H. 2006. The contribution ofvarious typesof activities to recovery.European JournalofWork andOrganizational Psychology, 15: 218–240.

Roth, L. M. 2003. Selling women short: A research note ongender differences in compensation on Wall Street.Social Forces, 82: 783–802.

Rothausen, T. J., Gonzalez, J. A., Clarke, N. E., & O’Dell,L. L. 1998. Family-friendly backlash—fact or fiction?The case of organizations’ on-site child care centers.Personnel Psychology, 51: 685–706.

Rothbard, N. P. 2001. Enriching or depleting? The dy-namics of engagement in work and family roles.Administrative Science Quarterly, 46: 655–684.

Rothbard, N. P., & Patil, S. V. 2012. Work engagement. InK. S. Cameron & G. M. Spreitzer (Eds.), The Oxfordhandbook of positive organizational scholarship:56–68. New York, NY: Oxford University Press.

Saxbe, D. E., Repetti, R. L., & Graesch, A. P. 2011. Timespent in housework and leisure: Links with parents’physiological recovery from work. Journal of FamilyPsychology, 25: 271–281.

Schaufeli, W. B., & Bakker, A. B. 2004. Job demands, jobresources, and their relationship with burnout andengagement: A multi-sample study. Journal of Orga-nizational Behavior, 25: 293–315.

Schaufeli, W. B., Bakker, A. B., & Salanova, M. 2006. Themeasurement of work engagement with a short ques-tionnaire a cross-national study. Educational andPsychological Measurement, 66: 701–716.

Schor, J. 1991. The overworked American: The unex-pected decline of leisure. New York, NY: Basic Books.

Schulz, J., & Watkins, M. 2007. The development of theleisure meanings inventory. Journal of Leisure Re-search, 39: 477–497.

SHRM Online Staff. 2011. Employees feel penalized forintegrating work, family. HRMagazine, 56: 20.


Smallwood, J. 2013. Distinguishing how from why themind wanders: A process–occurrence framework forself-generated mental activity. Psychological Bulle-tin, 139: 519–535.

Smit, B.W.,Maloney, P.W.,Maertz, C. P., &Montag-Smit,T. 2016. Out of sight, out of mind? How and whencognitive role transition episodes influence em-ployee performance. Human Relations, 69: 2141–2168.

Sonnentag, S. 2003. Recovery, work engagement, andproactive behavior: A new look at the interface be-tween nonwork and work. The Journal of AppliedPsychology, 88: 518–528.

Sonnentag, S., & Braun, I. 2013. Not always a sweethome: Family and job responsibilities constrain re-covery processes. In J. G. Grzywacz, E. Demerouti,J. G. Grzywacz & E. Demerouti (Eds.), New frontiersinwork and family research: 71–92. NewYork, NY:Psychology Press.

Sonnentag, S., & Fritz, C. 2007. The recovery experiencequestionnaire: Development and validation of a mea-sure for assessing recuperation and unwinding fromwork. Journal of Occupational Health Psychology,12: 204–221.

Sonnentag, S., & Grant, A. M. 2012. Doing good at workfeels good at home, but not right away:When andwhyperceived prosocial impact predicts positive affect.Personnel Psychology, 65: 495–530.

Sonnentag, S., & Zijlstra, F. R. H. 2006. Job characteristicsandoff-job activities aspredictors of need for recovery,well-being, and fatigue. The Journal of Applied Psy-chology, 91: 330–350.

Stryker, S. 1968. Identity salience and role performance:The relevance of symbolic interaction theory forfamily research. Journal ofMarriage and the Family,30: 558–564.

Ten Brummelhuis, L. L., & Bakker, A. B. 2012. Stayingengaged during the week: The effect of off-job activi-ties on next-day work engagement. Journal of Occu-pational Health Psychology, 17: 445–455.

Ten Brummelhuis, L. L., van der Lippe, T., & Kluwer, E. S.2010. Family involvement and helping behavior inteams. Journal of Management, 36: 1406–1431.

U.S. Bureau of Labor Statistics. 2014. Unpublishedmaritaland family tabulations (Current Population Survey)[Database]. Retrieved from www.bls.gov/cps

van Hooff, M. L. M., Geurts, S. A. E., Beckers, D. G. J., &Kompier,M.A. J. 2011. Daily recovery fromwork: Theroleof activities, effort andpleasure.WorkandStress,25: 55–74.

Vernon, V. 2010. Marriage: For love, for money. . . and fortime? Review of Economics of the Household, 8:433–457.

Viljevac, A., Cooper-Thomas, H. D., & Saks, A. M. 2012.An investigation into the validity of two measures ofwork engagement. International Journal of HumanResource Management, 23: 3692–3709.

Watson, D., Clark, L. A., & Tellegen, A. 1988. Developmentand validation of brief measures of positive and neg-ative affect: The PANAS scales. Journal of Personal-ity and Social Psychology, 54: 1063–1070.

Weinstein, N., & Ryan, R. M. 2010. When helping helps:Autonomousmotivation for prosocial behavior and itsinfluence on well-being for the helper and recipient.Journal of Personality and Social Psychology, 98:222–244.

Williams, J. C. 2001.Unbending gender: Why family andwork conflict and what to do about it. New York,NY: Oxford University Press.

Wilson, K. S., & Baumann, H. M. 2015. Capturing a morecomplete viewof employees’ lives outsideofwork:Theintroductionanddevelopmentofnew interrole conflictconstructs. Personnel Psychology, 68: 235–282.

Woolley, A. W. 2009. Means vs. ends: Implications ofprocess and outcome focus for team adaptation andperformance. Organization Science, 20: 500–515.

Zijlstra, F., & Cropley, M. 2006. Recovery after work. InF. Jones, R. J. Burke & M. Westman (Eds.). Work–lifebalance:Apsychologicalperspective: 220–234.NewYork, NY: Psychology Press.

Tracy L. Dumas ([email protected]) is an associateprofessor ofmanagement and human resources at theOhioState University’s Fisher College of Business. She earnedher PhD from Northwestern University. Her researchdraws primarily on identity, role, and boundary theoriesto consider how employees’ nonwork roles and identitiesshape their workplace experiences.

Jill E. Perry-Smith ([email protected]) is anassociate professor of organization and management atEmory University’s Goizueta Business School. She re-ceived her PhD from Georgia Institute of Technology. Herresearch focuses on a social network perspective ofcreativity. Her research also explores the effects of thework/nonwork interface on individual and organizationaloutcomes.


http://www.bls.gov/cps

mailto:[email protected]

mailto:[email protected]

APPENDIX A

After-Work Domestic and Leisure Activities Anticipated by Respondents

Study 2

Domestic Activity Count % Leisure Activity Count %

Childcare 52 17.99 Dinner with friends 30 34.09Miscellaneous 47 16.26 Miscellaneous 18 20.45Shopping/errands 36 12.46 Family/child outing 10 11.36Cooking 35 12.11 Cultural event 5 5.68Pet care 28 9.69 Sports/games 5 5.68Laundry 20 6.92 Drinks with friends 4 4.55Paperwork/taxes/bills 17 5.88 Shopping 4 4.55Dinner 14 4.84 TV 4 4.55Cleaning 12 4.15 Outing with friends 3 3.41Auto or home maintenance 11 3.81 Date 3 3.41Non-child caregiving 8 2.77 Movie 2 2.27Family time 6 2.08 Total Leisure 88Medical appointments 3 1.04Total Domestic 289

Study 2: Supplement

Domestic Activity Count % Leisure Activity Count %

Shopping/errands 58 20.49 Dinner with friends/dinner out 30 20.55Cooking 51 18.02 Sports/games 25 17.12Childcare 40 14.13 Family outing 23 15.75Miscellaneous 28 9.89 Miscellaneous 22 15.07Family time 27 9.54 Outing with friends 13 8.90Laundry 20 7.07 Drinks with friends/drinks out 9 6.16Cleaning 17 6.01 Shopping 6 4.11Pet care 15 5.30 TV 6 4.11Auto or home maintenance 14 4.95 Cultural event 4 2.74Medical appointments 5 1.77 Date 3 2.05Non-child caregiving 4 1.41 Movie 3 2.05Paperwork/taxes/bills 3 1.06 Child outing 2 1.37Dinner 1 0.35 Total Leisure 146Total Domestic 283


Copyright of Academy of Management Journal is the property of Academy of Managementand its content may not be copied or emailed to multiple sites or posted to a listserv withoutthe copyright holder's express written permission. However, users may print, download, oremail articles for individual use.

the paradox of family structure and plans after work: why single … · 2019. 4. 20. · between...

Documents