Time Management Plan
Work–Family Conflict and Self-Discrepant Time Allocation at Work
Patricia C. Dahm, Theresa M. Glomb, Colleen Flaherty Manchester, and Sophie Leroy University of Minnesota
We examine the relationships between work-to-family conflict, time allocation across work activities, and the outcomes of work satisfaction, well-being, and salary in the context of self-regulation and self-discrepancy theories. We posit work-to-family conflict is associated with self-discrepant time allocation such that employees with higher levels of work-to-family conflict are likely to allocate less time than preferred to work activities that require greater self-regulatory resources (e.g., tasks that are complex, or those with longer term goals that delay rewards and closure) and allocate more time than preferred to activities that demand fewer self-regulatory resources or are replenishing (e.g., those that provide closure or are prosocial). We suggest this self-discrepant time allocation (actual vs. preferred time allocation) is one mechanism by which work-to-family conflict leads to negative employee consequences (Allen, Herst, Bruck, & Sutton, 2000; Mesmer-Magnus & Viswesvaran, 2005). Using polynomial regression and response surface methodology, we find that discrepancies between actual and preferred time allocations to work activities negatively relate to work satisfaction, psychological well-being, and physical well-being. Self-discrepant time allocation mediates the relationship between work-to-family conflict and work satisfaction and well-being, while actual time allocation (rather than the discrepancy) mediates the relationship between work-to-family conflict and salary. We find that women are more likely than men to report self-discrepant time allocations as work-to-family conflict increases.
Keywords: self-discrepancy, work–family conflict, self-regulation
Time is the coin of your life. It is the only coin you have, and only you can determine how it will be spent. Be careful lest you let other people spend it for you.
—Carl Sandburg
Time, unlike other resources, is a universal constraint. As such, the finite nature of time has always held a key role in work–family research, suggesting time investment in either the work or family role affects time allocated to the other (Rothbard & Edwards, 2003). Accordingly, popular conceptions of work–family conflict consist of time, demands, strain, and behavior-based elements to assess the degree to which work responsibilities preclude fulfilling
family responsibilities or vice versa (e.g., Carlson, Kacmar, & Williams, 2000; Greenhaus & Beutell, 1985; Netemeyer, Boles, & McMurrian, 1996). Conflicting work and family demands have deleterious consequences for a wide range of outcomes including career success, job satisfaction, life satisfaction, somatic symp- toms, marital satisfaction, and family satisfaction (Allen, Herst, Bruck & Sutton, 2000; Mesmer-Magnus & Viswesvaran, 2005).
Whereas research has focused on perceptions of time strains between the work and family roles, it has been silent on the effects of work-to-family (WTF) conflict on time allocation within the work role. For example, the time facet of WTF conflict focuses on time incompatibilities between work and family domains. This interrole perception of time strains prevalent in work–family re- search can be complemented by an intrarole view of time at work, examining how WTF conflict affects the way individuals allocate their time to work activities and whether individuals achieve their time allocation preferences. Time allocation choices affect the ability to reach desired work goals, ultimately affecting chances for career success, work satisfaction, and well-being. Our focus is this intrarole time allocation at work.
We draw on self-regulation (e.g., Baumeister & Heatherton, 1996) and self-discrepancy (e.g., Higgins, 1987; Higgins, Bond, Klein, & Strauman, 1986) theories to develop theory for how WTF conflict relates to self-discrepant time allocation at work, which in turn affects work satisfaction, well-being, and salary. Acts of choice such as deliberations related to work and family demands can deplete self-regulatory resources (Baumeister & Vohs, 2007), as do the demands themselves. The depleted state makes it harder to allocate time to activities that require greater self-regulatory resources such as complex work activities or those with long-term goals and delayed rewards. When time allocation choices are out
This article was published Online First February 9, 2015. Patricia C. Dahm, Theresa M. Glomb, Colleen Flaherty Manchester, and
Sophie Leroy, Carlson School of Management, University of Minnesota. Sophie Leroy is now at the University of Washington Bothell School of
Business. Earlier versions of this article were presented at the 2012 Work and
Family Researchers Network Conference, New York, New York and the 2014 Society of Industrial Organizational Psychologists, Honolulu, Ha- waii. We are grateful to Lisa Leslie; Devasheesh Bhave; Joyce Bono; John Kammeyer-Mueller; the participants in Work and Organizations Collo- quium Series at the Carlson School of Management; and the participants in the Work, Family, and Time Seminar Series at the Minnesota Population Center for helpful comments on earlier versions of this article. We thank the participants for their survey responses and the organization for provid- ing access to their database.
Correspondence concerning this article should be addressed to Patricia C. Dahm, Carlson School of Management, University of Minnesota, 321 19th Avenue South, Suite 3-300, Minneapolis, MN 55455. E-mail: dahmx021@umn.edu
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
Journal of Applied Psychology © 2015 American Psychological Association 2015, Vol. 100, No. 3, 767–792 0021-9010/15/$12.00 http://dx.doi.org/10.1037/a0038542
767
of alignment with one’s preferences, it causes self-discrepancy (Higgins et al., 1986).
We propose self-discrepant time allocation as a key mediator between WTF conflict and well-being and career outcomes. Re- search examining explanatory mechanisms for the effects of WTF conflict has focused largely on affective and stress-related mech- anisms such as guilt and hostility (Judge, Ilies, & Scott, 2006), emotional strain (Rothbard, 2001), quality of work and family life (Duxbury & Higgins, 1991), or work and family distress (e.g., Frone, Russell, & Cooper, 1992; Grandey & Cropanzano, 1999) rather than potential changes in employee work behaviors. In- creased attention to explanatory mechanisms that include on-the- job behavior such as intrarole time allocation is needed.
Although work–family conflict is bidirectional in nature, our focus is on WTF conflict. When individuals are faced with con- flicting work and family demands, the direction of the conflict (WTF or family-to-work) “is only apparent after an individual decides to participate in one or the other activity” (Greenhaus & Powell, 2003, p. 292; see also Greenhaus & Beutell, 1985). So a parent choosing to go to a work meeting instead of a school event results in WTF conflict; whereas choosing to attend the school event instead of the meeting results in family-to-work conflict. Research has shown that WTF, as compared to family-to-work conflict has a stronger relationship with work outcomes including job satisfaction, work satisfaction, intention to turnover, and work- related stress (e.g., Amstad, Meier, Fasel, Elfering, & Semmer, 2011; Michel, Mitchelsen, Kotrba, LeBreton, & Baltes, 2009; Shockley & Singla, 2011).
Scholars have explained WTF conflict’s stronger relationship to work outcomes using source attribution theory, which proposed that “negative affective reactions (and their possible behavioral consequences, such as quitting) are likely to center around the domain that is seen as causing the problem” (Amstad et al., 2011, p. 153) because individuals blame work as the source of the conflict. Also, consistently choosing work over family demands may make resource recovery less likely, making WTF conflict more depleting than choosing family over work (Demerouti, Bak- ker, & Bulters, 2004; Demerouti, Bakker, & Schaufeli, 2005). Research has shown that as compared to family-to-work conflict, WTF conflict has a stronger effect on burnout and exhaustion (Amstad et al., 2011; Reichl, Leiter & Spinath, 2014).
We make manifold contributions to the work–family literature. First, we take an innovative intrarole view of time to complement perceptions of interrole time strain predominant in work–family conflict measures. Second, we apply self-regulation theory to demonstrate how WTF conflict may generate discrepancies be- tween employees’ actual and preferred time allocations such that less time than preferred is allocated to work activities requiring greater resources in favor of those activities that demand fewer self-regulatory resources or are replenishing. Third, applying self- discrepancy theory (Higgins et al., 1986), we posit differences between actual and preferred time allocations adversely affect work satisfaction, and psychological and physical well-being, and that actual time allocation affects salary. We elucidate self- discrepant time allocation as a novel mediating mechanism through which WTF conflict affects well-being and career out- comes, thereby providing needed attention to work-related behav- ioral mediating mechanisms and to the relationship between WTF conflict and objective career success (i.e., salary).
Derivation of Hypotheses
Self-Discrepant Time Allocations at Work
Self-discrepancy theory suggests that the self-concept allows for multiple self-representations (Higgins, 1987; Higgins et al., 1986), which include an actual self (who one is), but may also include past selves (who one once was; Albert, 1977), possible selves (who one might be; Markus & Nurius, 1986), ideal selves (who one would like to be; Higgins, 1987), ought selves (who one should be; Higgins, 1987), and alternative selves (who one could have been; Obodaru, 2012). When a person’s hopes, goals, or wishes that make up their ideal self (Markus & Nurius, 1986) remain unful- filled, individuals experience differences between the person one is (actual self) and the person one wants to be (ideal self), which is referred to as self-discrepancy (Higgins, 1987; Tangney, Nie- denthal, Covert, & Barlow, 1998). Differences between actual and ideal selves challenge an individual’s sense of self (Pratt, 2000) and create an uncomfortable “identity deficit,” which individuals are motivated to resolve (Pratt & Dutton, 2000). Despite motiva- tion to attain ideal self-states, individuals are often unable to, leading to feelings of dejection (e.g., dissatisfaction, discourage- ment, sadness, depression, low self-esteem, shame) and agitation (e.g., anxiety, worry, fear, spells of terror, or panic; Higgins, 1987; Higgins, Klein, & Strauman, 1985; Moretti & Higgins, 1990; Tangney, Niedenthal, Covert, & Barlow, 1998).
In this paper, we focus on discrepancies in the work domain between the actual self, including attributes one actually has and behaviors one actually exhibits, and the ideal self, including attri- butes one would like to have and behaviors one would like to exhibit. Self-discrepancy is conceptualized as differences between actual time allocation to categories of work activities (i.e., the actual self) and preferred time allocations (i.e., the ideal self). For example, a sales professional and a faculty member have prefer- ences for how to allocate their time to achieve goals and ideals (ideal self), but actual time allocation (actual self) may not align with these preferences. The salesperson may prefer to allocate most of her time to building new customer relationships and maintaining current ones, which is consistent with her goals for rewards and advancement, but actually spend much of her time on administrative paperwork. Likewise, a faculty member at a re- search institution may prefer to spend the majority of her time on research, but instead, spend more time than preferred on teaching or service. Time allocation preferences should reflect individu- als’ goals because “all motivated individuals pursue goals” (Sheldon & Elliot, 1999, p. 482), goal progress generates pos- itive affect (Carver & Scheier, 1990), and goal attainment improves well-being (Brunstein, 1993; Elliott & Sheldon, 1997; Elliot, Sheldon, & Church, 1997; Sheldon & Kasser, 1998). Further, because organizations cultivate norms and create in- centives that define career success, an individual’s goals and ideals are likely to align with those of the organization. There- fore, misalignment of actual and preferred time allocation will not only create self-discrepancy and result in the negative psychological well-being outcomes associated with this aver- sive state (e.g., dissatisfaction and anxiety), but will also be detrimental to career success.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
768 DAHM, GLOMB, MANCHESTER, AND LEROY
WTF Conflict and Self-Discrepant Time Allocations at Work
WTF conflict and depletion. Work–family research has tra- ditionally taken the perspective that conflict between work and family roles stems from a lack of sufficient resources (time and energy) to meet demands (e.g., Greenhaus & Buetell, 1985). This conflict depletes resources; Grandey and Cropanzano (1999) ap- plied the Conservation of Resources Theory (Hobfoll, 1989) to suggest that interrole conflict leads to stress because resources (i.e., time and energy) are lost in the process of juggling work and family roles. Theory and meta-analytic results suggest WTF con- flict may be particularly depleting because work interference with family inhibits opportunity for resource recovery because
the person’s psychobiological system remains activated and does not stabilize at baseline level” and “the person has to make additional (compensatory) effort to maintain his or her level of performance, which leads to extra psycho-physiological costs that, in turn, interfere with the recovery process. (Demerouti et al., 2005; see also Demerouti et al., 2004).
Meta-analytic estimates show that work-to-nonwork conflict more strongly relates to burnout and exhaustion (r � .61) as compared to nonwork-to-work (r � .34; Reichl et al., 2014). This may be because family interactions have greater potential for positive effects on well-being (Bono, Glomb, Shen, Kim, & Koch, 2013; Gable, Reis, Impett, & Asher, 2004; Heaphy & Dutton, 2008). Although work interactions may have benefits such as reducing exhaustion (Halbesleben, 2006) spending more time at work (at the expense of one’s family time) has detrimental health effects (Galinsky, Kim, & Bond, 2001). Therefore, we focus on WTF conflict as the more consistently depleting effect.
To understand how resource loss from WTF conflict affects work behaviors, we employ the lens of self-regulation theory. Self-regulatory resources can be thought of as strength or energy, and are necessary for the self to exert control—that is, to be able to override incipient (i.e., initial) responses and replace them with those more consistent with “one’s long-range goals, ideals, re- solves or plans” (Baumeister, 2002, p. 670). Self-regulatory re- sources aid in the attainment of long-term goals and desired outcomes. Acts requiring self-regulatory resources include control- ling thoughts or emotions, regulating impulses and resisting temp- tations, persisting at difficult tasks, sustaining attention, and mak- ing decisions and choices; individuals have limited self-regulatory resources, which are depleted when engaging in such activities (Baumeister, Bratslavsky, Muraven, & Tice, 1998; Baumeister, Heatherton, & Tice, 1994; Baumeister & Vohs, 2007; Hagger, Wood, Stiff, & Chatzisarantis, 2010; Vohs et al., 2008).
Theoretically, experiencing WTF conflict may deplete self- regulatory resources via three pathways. First, WTF conflict in- volves experiencing conflicting demands on one’s time and en- ergy, which requires people to deliberate, ascribe priorities, and make choices about these demands (Greenhaus & Powell, 2003; Powell & Greenhaus, 2006). Acts of deliberation and choice, even when easy, lead to self-regulatory depletion (Baumeister & Vohs, 2007). Second, in resolving competing work and family demands, people often sacrifice one goal for another, with one goal being postponed or not fully achieved. Such delay of gratification re- quires self-regulatory resources (Schmeichel & Inzlicht, 2013) as
does suppressing desires, thereby depleting self-regulatory re- sources (Baumeister et al., 1998). Last, experiencing conflicting work and family demands is associated with distress (Frone et al., 1992). Experiencing distress undermines individuals’ subsequent self-control (Baumeister et al., 1994; Tice, Bratslavsky, & Baumeister, 2001), demonstrating a depleting effect on self- regulatory resources. In summary, WTF conflict is likely to de- plete self-regulatory resources through deliberation and choice, suppressing or delaying desires, and distress.
When WTF conflict depletes self-regulatory resources, activities requiring self-regulation become more difficult and more likely to fail (Baumeister et al., 1998; Muraven & Baumeister, 2000). For example, depletion of self-regulatory resources is linked to subse- quent overeating by dieters, prejudicial responding, inappropriate sexual responses, and impulsive overspending (Gailliot & Baumeister, 2007; Richeson & Shelton, 2003; Vohs & Faber, 2007; Vohs & Heatherton, 2000). When depleted, people tend to want to feel better, choosing behaviors that offer more immediate gratification (Baumeister et al., 1994; Tice et al., 2001). Simply put, depletion of self-regulatory resources pivots people from activities that are more beneficial (but more depleting) to those that are more rewarding or replenishing in the short term, even if this diverts them from their long-term desired goals or preferences (Schmeichel & Inzlicht, 2013). For example, dieters with depleted self-regulatory resources are more likely to overeat and choose food that is gratifying in the present but undermines their long- term goal of weight loss (Vohs & Heatherton, 2000). Likewise, depleted individuals are more likely to engage in unethical behav- ior, despite goals for ethical behavior (Gino, Schweitzer, Mead, & Ariely, 2011). Parallels exist at work; just as dieters know over- eating hampers their long term goals, employees know excessive time responding to e-mail instead of working on a difficult or long-range project hampers their long term goals, and yet they may make such time allocations when depleted.
WTF conflict and time allocation. Time allocation choices may be either conscious or nonconscious (see Ryan & Deci, 2006); research has suggested both intuitive and deliberate processes guide choices. Deliberate monitoring systems operate to ensure choices are consistent with one’s goals or external standards (e.g., organizational, social, moral norms; Pocheptsova, Amir, Dhar, & Baumeister, 2009). When depleted, the monitoring system is not engaged or is more easily overridden. Once again, the example of dieters who choose unhealthy food over healthy options when depleted (e.g., Muraven & Baumeister, 2000) is illuminating. Even though individuals are conscious of the unhealthy choice they made, they may not be conscious of the reasons for this choice. This is consistent with research indicating even highly complex behaviors may be performed without conscious awareness of the “triggers” and processes guiding behavior (Aarts, 2007; Bargh, 1990; James, 1890; Papies & Aarts, 2011). Likewise, goals may be activated and pursued without conscious awareness (Chartrand & Bargh, 1996).
Unlike much of the self-regulation research that examined one specific behavioral response, allocating time at work requires consideration of multiple activities. Thus, the zero-sum nature of time results in an inherent connection between self-discrepant time allocation for any particular work activity and allocation of time for all other work activities. Holding total time fixed, there would be perfect alignment of actual and preferred time allocation if the
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
769WORK–FAMILY CONFLICT AND TIME ALLOCATION
individual only faced one activity. However, in most work settings individuals must allocate time between two or more work activi- ties, so time spent on one activity precludes time on any other activity. Using the example of academic faculty, holding work hours constant, spending more time on service than preferred will necessarily result in spending less time than preferred to research and/or teaching.
Research has provided insight on the types of activities individ- uals might favor or eschew as WTF conflict increases and moni- toring systems are more easily overridden. We propose that indi- viduals experiencing the depleting effects of WTF conflict will find it more difficult to allocate time according to their ideals. In particular, they will allocate less time than preferred to work activities that are further depleting including (1) complex tasks, or (2) those with long-term goals, which delay rewards or closure (e.g., Vohs et al., 2008), and will spend more time than preferred on work activities that are not further depleting (and may actually be replenishing) such as those providing more immediate closure or reward. Complexity can be characterized by path-goal multi- plicity, which increases with both the unknown number of paths to a goal, and uncertainty related to the best path to a goal (Campbell, 1988; March & Simon, 1958; Terborg & Miller, 1978). When depleted, individuals are not motivated to persist in complex tasks because they require cognitive processing, which is further deplet- ing (e.g., Hagger et al., 2010). Rewards also may be more delayed and less certain given the multiple and uncertain paths to goal accomplishment. Experiencing ambiguity is distressing; individu- als prefer tasks that provide cognitive closure, defined as “an epistemic state of cognitive clarity and certainty” (Webster, Rich- ter, & Kruglanski, 1996, p. 182; see also Heaton & Kruglanski, 1991; Kruglanski, 1989, 1990). Likewise, lack of completion is cognitively and emotionally aversive (e.g., Leroy, 2009), and activities that require delayed gratification or prolonged efforts toward a goal will be more difficult when WTF conflict and depletion are high. Therefore, individuals experiencing WTF con- flict will likely allocate less time than preferred to complex activ- ities or those with long-term goals and delayed rewards and more time than preferred to those that minimize cognitive processing or provide a sense of closure at work.
When WTF conflict is high, employees also may allocate more time than preferred to replenishing activities such as those with prosocial benefits because helping leads to improved mood (Glomb, Bhave, Miner, & Wall, 2011) and increased self-efficacy and motivation (Grant et al., 2007; Grant & Gino, 2010). Depleted employees also may have a more difficult time declining the requests of others, in an effort to gain positive reinforcement or to avoid feeling badly for declining; negative affect has been asso- ciated with ego depletion (Hagger et al., 2010). Despite long-term goals, individuals experiencing the depleting effects of WTF con- flict may be inclined toward tasks that provide a sense of closure or prosocial benefits, as these tasks would not further deplete regulatory resources and may even replenish them.
Hypothesis 1 (H1): WTF conflict will be related to self- discrepant time allocation such that time allocation will be- come more discrepant from ideals as WTF conflict increases. Specifically, as WTF conflict increases, actual time allocated to work activities that require self-regulatory resources will be less than preferred, and actual time allocated to work activities
that require fewer self-regulatory resources or are replenishing will be more than preferred.
Though we expect WTF conflict to affect actual time allocation, we do not expect WTF conflict to affect preferred time allocation for two reasons. First, ideal self-states are more stable than actual self-states (Strauman, 1996). Second, preferred time allocation is likely to incorporate organizational and individual work goals, which are longer term and more stable.
Self-Discrepant Time Allocation, Work Satisfaction, and Well-Being
Self-discrepancy theory proposes self-discrepant states are re- lated to both dejection and agitation related outcomes (e.g., shame, depression, feelings of worthlessness, panic, lower self-esteem; Higgins, 1987; Higgins et al., 1985; Moretti & Higgins, 1990; Polasky & Holahan, 1998; Tangney et al., 1998). For example, failing to spend time developing new sales leads, despite inten- tions, may cause a sales professional dissatisfaction about her job and anxiety about meeting a sales goal, which brings with it career success. A faculty member who spends less time on research than intended may worry about her record for tenure, bringing dissat- isfaction and threats to her work identity. We posit the dejection and agitation outcomes suggested in self-discrepancy theory may be represented by well-being and career outcomes including: work satisfaction, psychological well-being, and physical well-being.
Self-discrepancy will have unfavorable effects through its threats to (1) goal attainment; (2) the meaning, fulfillment, and enjoyment gained from working; and (3) one’s work identity. Allocating less time than preferred to activities that support work goals threatens goal attainment; progress toward goals is positively related to job satisfaction and well-being (Judge, Bono, Erez, & Locke, 2005; Koestner, Lekes, Powers, & Chicoine, 2002; Maier & Brunstein, 2001). Lack of goal attainment, related to both chronic and more immediate discrepancies, is associated with agitation and tension (Higgins, Shah, & Friedman, 1997). Time allocation preferences also reflect preferences for certain job char- acteristics, such as complexity or social interactions (Morgeson & Humphrey, 2006). Therefore, self-discrepant time allocation im- pedes fulfillment of those characteristics and the consequent mean- ing, fulfillment, or enjoyment experienced at work. Finally, if time allocation preferences reflect one’s ideal work identity, individuals who fail to achieve their ideal self-states experience identity def- icits and the consequent anxiety. For example, individuals expe- rience discomfort when they realize they are not as wealthy as they would like to be (Pratt, 2000) and anxiety when they behave differently from the mother they would like to be (Polasky & Holahan, 1998). Lack of goal attainment and identity deficits associated with self-discrepant time allocation create anxiety, which is detrimental to health. Given individuals spend more time working than in any other primary life activity (American Time Use Survey; Bureau of Labor Statistics, 2010), self-discrepant time allocation is likely to affect personal well-being; a link between goal attainment and life satisfaction has been demonstrated (Judge et al., 2005).
The relationship between the discrepancy and outcomes may take three functional forms: asymptotic, monotonic, or U-shaped (Shockley, 2013). All three functional forms assume that as actual
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
770 DAHM, GLOMB, MANCHESTER, AND LEROY
time allocation falls short of preferred, employees will report worse outcomes. However, these functional forms differ in their effect on outcomes as actual time allocation exceeds preferences. An asymptotic relationship suggests that once actual time alloca- tion exceeds preferred time allocation, excess time allocation can neither improve nor detract from well-being or career success. A monotonic relationship suggests that as actual time allocation exceeds preferred time allocation, employee outcomes improve. A U-shaped relationship suggests that as actual time allocation ex- ceeds preferred, outcomes would decline, in the same way that they do when actual falls short of preferred; congruence results in the most favorable outcomes.
For work satisfaction and well-being, we hypothesize the rela- tionship between actual and preferred time allocation will be U-shaped, such that allocating either more or less time than pre- ferred will result in negative outcomes; congruence is optimal. This is consistent with self-discrepancy theory, which holds the magnitude of the discrepancy is important, but does not distinguish between falling short from or exceeding preferences.
Hypothesis 2: The discrepancy between actual and preferred time allocation to work activities is associated with: (H2a) work satisfaction, (H2b) psychological well-being, and (H2c) physical well-being. These relationships take the form of an inverted U-shaped curve, such that outcomes decrease as actual time allocation becomes discrepant from preferred.
Self-Discrepancy and Salary
Meta-analytic evidence suggests work hours are only weakly associated with higher salary and do not significantly relate to promotions (Judge, Cable, Boudreau, & Bretz, 1995), suggesting perhaps the number of hours people work may not be as important as what they do during those hours. In the case of salary, individ- uals are likely to be rewarded for time allocation to work activities that require self-regulatory resources (regardless of preferences for such activities) because they are more complex, take longer to complete, and are typically valued by the organization. For exam- ple, sales professionals who spend time cultivating valued new sales leads are more likely to make more sales and to be compen- sated in the form of commission. In academia, time spent on valued research activities is likely to translate into publications, and publications— both quantity and quality—translate into salary (e.g., Bellas & Toutkoushian, 1999; Persell, 1983). Allocating time to more complex work activities with delayed rewards and feed- back is likely to be associated with increased salary even beyond one’s preferred allocation (i.e., more is better). Though individuals may become less satisfied with their time allocations, they may still be rewarded for them. Therefore, in contrast to the U-shaped relationship proposed for satisfaction and well-being outcomes, we expect a linear relationship between salary and time allocation to work activities requiring self-regulatory resources.
Hypothesis 3: Salary will increase as actual time allocation to work activities that require self-regulatory resources increases, regardless of time allocation preferences.
Mediation Hypotheses
Research has suggested a negative relationship between WTF conflict and the work satisfaction and well-being outcomes spec-
ified in H2 (i.e., Allen et al., 2000; Mesmer-Magnus & Viswes- varan, 2005). Though there is evidence that emotions and job distress mediate the relationship between WTF conflict and out- comes (Frone et al., 1992; Grandey & Cropanzano, 1999; Judge et al., 2006; Rothbard, 2001), explanatory mechanisms related to work behaviors remain underdeveloped. Following from H1, pro- posing WTF conflict relates to self-discrepant time allocation, and H2, proposing self-discrepant time allocation negatively affects work satisfaction and well-being outcomes, we propose self- discrepant time allocation as an explanatory mediating mechanism for the relationship between WTF conflict and work satisfaction, psychological well-being, and physical well-being.
Hypothesis 4: Self-discrepant time allocation will mediate the effects of WTF conflict on (a) work satisfaction, (b) psycho- logical well-being, and (c) physical well-being.
Although the relationship between WTF conflict and work satisfaction and well-being is well documented, the relationship between WTF conflict and salary is less clear. On the one hand, allowing work to interfere with family may indicate dedication to work, for which individuals may be rewarded. However, our theory suggests WTF conflict relates to difficulty allocating time to activities requiring self-regulatory resources, such as those that are complex or offer delayed gratification, which are more likely to be valued and rewarded by the organization. Hefty cognitive demands are among the strongest predictors of wages (e.g., Bound & Johnson, 1992; Juhn, 1999), and executive compensation is a function of job complexity (Agarwal, 1981). If WTF conflict relates to individuals spending less time on tasks that require regulatory resources, but these tasks are the highly rewarded tasks, then the effects of WTF conflict on salary through time allocation will be detrimental. Meta-analysis has shown a small, negative relationship between WTF conflict and job performance (Allen et al., 2000), which is one step removed but predictive of salary (e.g., Bishop, 1987); this relationship also has the potential to be ex- plained by not allocating sufficient time to tasks requiring self- regulatory resources.
Following from H1, which proposed a negative relationship between WTF conflict and time allocation to activities requiring self-regulatory resources; and H3, which proposed such time al- location relates to salary, we propose actual time allocation to activities requiring self-regulatory resources (rather than the dis- crepancy) will mediate the relationship between WTF conflict and salary.
Hypothesis 5: Time allocated to work activities that require self-regulatory resources will mediate the effects of WTF conflict on salary.
Self-Discrepancy and Gender
Given that work–family conflict is often studied through the lens of gender role theory (cf. Grandey, Cordeiro, & Crouter, 2005), and there are known, but largely unexplained, gender dif- ferences in career success (e.g., Blau & Kahn, 2006), we investi- gate the role of gender in these relationships. Theoretically, gender may operate on the relationship between WTF conflict and time allocation because of gender differences in social role expectations such that, as compared to men, (1) WTF conflict may be more
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
771WORK–FAMILY CONFLICT AND TIME ALLOCATION
depleting for women; and (2) once depleted, women will have a tougher time adhering to their own preferences.
Social role theory (Eagly, 1987) suggests that women are ex- pected to be more communal (i.e., concerned with the welfare of others, including being helpful, kind and interpersonally sensitive), and men are expected to be more agentic (i.e., more self-serving, including being assertive, ambitious, and dominant). Social role theory extends to family roles suggesting being a good wife and mother should take priority over work pursuits for women; more women than men attribute greater importance to the family role than the work role, while more men than women attribute more importance to the work role (Cinamon & Rich, 2002). Evidence suggests men and women incorporate social role expectations into their self- concepts such that they affect their ideal and ought selves (W. Wood, Christensen, Hebl, & Rothgerber, 1997). Further, men and women are rewarded for behaving in accordance with expecta- tions; many studies detail the interpersonal penalties for women when they violate social role expectations (e.g., Heilman & Okimoto, 2007; Heilman, Wallen, Fuchs, & Tamkins, 2004). So- cial role expectations suggest women will receive more role pres- sure from others to choose family over work, and “strong role pressure arouses a more intense force to comply within the indi- vidual than does a weak role pressure because a strong pressure implies more substantial sanctions for compliance or noncompliance” (Greenhaus & Powell, 2003, p. 292; see also Kahn et al., 1964). Indeed, women have more difficulty than men making the trade-off between family and work obligations (Bolino & Turnley, 2005; Tenbrunsel et al., 1995). Thus, WTF conflict presents a counternor- mative and more depleting choice for women, compared to men.
Once depleted, self-discrepant work time allocation may be more likely for women because social role expectations work in favor of self-concordant time allocation for men, but against self-concordant time allocation for women. Men will be rewarded for allocating time in accordance with their own ideals because this is agentic behavior, whereas women may be penalized for allocating time in accordance with their own ideals because it may be viewed as self-serving. Indeed, women, more so than men, incorporate the standards of others into their self-construction and behave in accordance with standards others hold for the self. As noted by Moretti and Higgins (1999), women experience more distress than men when they perceive their actual selves as discrepant from standards that others hold for them and “may seek congruency with others’ standards at the price of discrepancy with their own standards for the self” (p. 208). Overall, women may require more self-regulatory resources to allocate time in accordance with their own ideals, making self-discrepant time allo- cation more likely when WTF conflict is high.
Hypothesis 6: WTF conflict will relate to self-discrepant time allocation more strongly for women as compared to men.
Method
Participants and Procedure
For our primary analysis, we e-mailed 3,834 survey invitations to faculty at a large public university in the spring of 2010 and received 1,503 completed surveys for a response rate of 39%. This response rate compares favorably with the wider organizational literature (Roth & BeVier, 1998) and is very good for this partic-
ular employee group in this organization. Using a unique identi- fier, we linked this survey data with administrative records that identified each participant’s salary and academic rank.
We limited the sample in two ways. First, we restricted our sample to the 1,367 tenured and tenure-track faculty because work activities related to the troika of research, teaching, and service are fairly uniformly expected in this group, as compared to instructional and adjunct faculty who primarily teach. Second, we restricted the sample to respondents whose actual and preferred time allocations each totaled 100% (see Measures below), for a total of 1,243 faculty members. Eighty-three percent of participants were White, 40% were women, average organizational tenure was 16 years, and average age was 51 years. The distribution of academic rank was: 351 (28%) assistant, 368 (30%) associate, and 524 (42%) full professors.
In many ways, a faculty sample is ideal to test ideas of time allocation as faculty are used to categorizing work into research, teaching, and service activities. Although there can be synergies among these categories (e.g., reviewing a paper as a service activity helps a research project, research findings foster classroom content), faculty activities typically fall into one category. In addition, faculty members have high job control (a perceived ability to exert some influence over one’s work environment; Ganster, 1989) and have discretion over time allocation.
To strengthen inferences from our primary analyses, we surveyed the same population of employees approximately 2 years after (Time 2) our original survey (Time 1), and conducted longitudinal analyses on a subset of our sample that participated in both surveys (N � 595). All study variables were collected at both Time 1 and Time 2. In the Time 2 sample, 83% of participants were White, 42% were women, average organizational tenure was 17.6 years, and average age was 53 years. The distribution of academic rank was: 120 (20%) assistant, 210 (35%) associate, and 265 (45%) full professors.
Measures
Time allocations. Time allocation measures were adapted from the National Study of Postsecondary Faculty (NSOPF) and Winslow (2010). To measure actual time allocation, participants were asked: “Please indicate what percent of your work time you spend on teaching, research, service, and administrative duties during a typical semester at your academic job. Please ensure that your indicated percentages total 100%.” To measure preferred time allocation, participants were asked a parallel item: “Please indicate what percent of your work time you would prefer to spend on teaching, research, service, and administrative duties during a typical semester at your academic job.” Descriptions of teaching, research, service, and administration were provided.1
Self-discrepancy. To operationalize self-discrepant time allo- cations to work activities that require relatively higher or lower self-regulatory resources, we must first determine the types of work activities most commonly mapped to those categories. Within the faculty context, we posit that of the categories of research, teaching, and service, the research domain most consis- tently includes tasks that are complex, relevant to longer term goals, with delayed rewards and closure, and thus require more
1 Because not all faculty have an administrative appointment, but all have research, teaching, and service obligations, we did not address time allocation to administrative duties in our analysis.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
772 DAHM, GLOMB, MANCHESTER, AND LEROY
self-regulatory resources. Teaching and service activities are more likely to provide more proximal closure, and to be prosocial thereby requiring fewer self-regulatory resources, and having the potential to be replenishing. To support the above propositions, we conducted a small survey to assess how research, teaching, and service activities vary along dimensions relevant to self-regulatory resources including complexity, proximity of goals/delayed grati- fication, task closure, and prosocial characteristics. Our sample included 28 faculty members at research universities (92% were White, 59% were women, average age � 42.4, 22% were assistant professors, 33% were associate professors, and 44% were full professors). We measured complexity with two items adapted from the job complexity scale (Humphrey, Nahrgang & Morgeson, 2007), proximity of goals/delayed gratification with two items adapted from the definition of delayed gratification provided in (Mischel, Cantor, & Feldman, 1996), task closure with one item adapted from the task identity scale (Humphrey et al., 2007), and prosocial characteristics with two items adapted from the prosocial job characteristics scale (Grant, 2008). Comparing mean differ- ences, we find research activities are reported to have significantly more complexity, longer term goals/delayed gratification, but less task closure and prosocial characteristics as compared to service and teaching (see Table 1, Figure 1). We also assessed negative affect related to declining research, teaching, or service activities with one item (e.g., “I feel bad saying ‘no’ to service-related requests from others”), and found individuals did not feel signif- icantly worse when declining service-related requests, as com- pared to research and teaching. In general, these data support our categorization of research activities as requiring higher self- regulatory resources than teaching and service.
Self-discrepant time allocation is operationalized in three ways to evaluate our hypotheses. First, domain-specific (i.e., research, teaching, and service) self-discrepant time allocations were calcu- lated as the difference between the preferred and actual percentage time allocations in each domain (i.e., preferred � actual). Al- though not common in other contexts, difference scores have been a preferred and valid operationalization of self-discrepancy (cf. Liss, Schiffrin, & Rizzo, 2013; Tangney et al., 1998; Winslow, 2010). A positive value indicates that actual time allocation is less than preferred, and a negative value indicates that actual time allocation is more than preferred. This difference score was used to estimate the magnitude and direction of the relationship between WTF conflict and self-discrepant time allocations (as a dependent or mediating variable as in H1 and H4). Second, the total self- discrepant time allocation was calculated as the sum of the abso-
lute value of the discrepancies in each of the research, teaching, and service domains (as a dependent or mediating variable as in H1 and H4). Finally, we used a quadratic in actual time allocation and preferred time allocation, plus the interaction between actual and preferred time allocations, for our tests of relationships be- tween self-discrepancy and outcomes (H2 and H3) in polynomial regressions and response surface methodology as recommended by Edwards and colleagues when discrepancy (or congruence) effects are the independent variables (e.g., Edwards & Parry, 1993).
WTF conflict. WTF conflict was assessed with five items from the work-family conflict scale (Netemeyer et al., 1996) on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree; � � .95). A sample item is “The demands from work interfere with my home and family life.”
Work satisfaction. Work satisfaction was measured with six items (� � .77) including five items from the Satisfaction with Work facet of the Job Descriptive Index (JDI; Smith, Kendall, & Hulin, 1969) as abridged by AJDI (Stanton et al., 2001). Partici- pants indicated the extent to which the following described their work: “sense of accomplishment, dull, satisfying, uninteresting, challenging” with the scale 3 (yes), 1 (not sure) to 0 (no; � � .75). Work satisfaction also included one item assessing global satis- faction: “Overall, I am satisfied with my employment at the university,” with the following scale ranging from 1 (strongly disagree) to 5 (strongly agree; modified from Wanous & Hudy, 2001). Items were standardized and combined.
Table 1 Properties of Research, Service, and Teaching
Research Service Teaching Research–Service
(Difference) Research–Teaching
(Difference)
M (SD) M (SD) M (SD) M (SD) M (SD)
Complexity 4.87 0.33 3.09 0.99 3.54 0.75 1.80�� 1.03 1.33�� 0.83 Delayed gratification 4.67 0.42 2.77 0.74 2.85 0.85 1.87�� 0.95 1.81�� 1.02 Closure 2.63 0.97 3.39 1.17 4.00 1.04 –0.74�� 1.83 –1.37�� 1.62 Prosocial 2.65 0.88 3.54 0.79 4.41 0.50 –0.87�� 1.17 –1.76�� 1.20 Feel bad saying “no” 3.52 0.98 3.68. 0.86 3.33 0.96 –0.22 1.19 0.19 1.55
�� p � .01.
0
1
2
3
4
5
Research (M)
Service (M)
Teaching (M)
Figure 1. Self-regulatory properties of research, service, and teaching. Items were measured on a scale of 1 (strongly disagree) to 5 (strongly agree). See the online article for the color version of this figure.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
773WORK–FAMILY CONFLICT AND TIME ALLOCATION
Psychological well-being. Psychological well-being was mea- sured with two items (� � .61). The first item: “During the past six months, how often have you felt nervous and stressed?”, was scored with a scale ranging from 1 (never) to 6 (very often; reverse scored) and was modified from the General Health Questionnaire (Goldberg & Hillier, 1979). The second item was: “In general, I am satisfied with my life,” and was scored with a scale ranging from 1 (strongly disagree) to 5 (strongly agree; Diener, Emmons, Larsen, & Griffin, 1985). Items were standardized and combined.
Physical well-being. To assess physical well-being, participants were asked: “In general, my health is” on a 5-point scale ranging from 1 (poor), 2 (fair), 3 (good), 4 (very good), to 5 (excellent). The item is adapted from DeSalvo and colleagues (2006) who found that single-item measures of general health demonstrated good reliability and strong concurrent validity with established multi-item health measures.
Annual salary. Annual salary was taken from administrative records and was merged with survey data. To account for the non- normal distribution, we used the natural log of salary in our analyses.
Controls. Controls for both work related and personal factors were included. Work-related controls were academic rank, hours worked, and administrative duties, as these variables have relevance for time allocation. Participants indicated their academic rank (assis- tant professor � 2, associate professor � 3, full professor � 4), and we created dummy variables for assistant professor and associate professor status. Hours worked were obtained by asking: “How many hours do you work in a typical week?” Administrative duties were assessed with: “At present, do you have an administrative appoint- ment?” (1 � yes, 0 � no).
In the personal domain, we included age, parental status, marital status, spouse employment status, and gender controls given their relevance to the experience of work and family demands. Participants indicated their gender as: female, male, transgender, other, or choose not to answer. A dummy variable was created for gender (1 � female; 0 � male; transgender, other, and choose not to answer were coded as missing). Participants indicated the number of children or dependents age 18 years or younger (0 – 7�). A dummy variable was created to indicate parental status (1 � at least one dependent age 18 or younger; 0 � no dependents age 18 or younger), an approach con- sistent with past literature (e.g., Leslie, Manchester, Park, & Mehng, 2012). Participants reported marital status using the options: single, married, same-sex domestic partner, living with a significant other or partner, divorced or separated, widowed, or choose not to answer. Participants were then asked, “If you are married, have a same-sex domestic partner, or are living with a partner, is your spouse or partner employed?” (yes, no, or choose not to answer). We created a dummy variable for married/partnered and spouse employed (1 � married, same-sex domestic partner, or living with a significant other or partner who is employed). We created a second dummy variable for married/partnered and spouse not employed (1 � married, same-sex domestic partner, or living with a significant other or partner who is not employed). The reference group for these dummy variables is those who were single, divorced or separated, or widowed.
Analysis
H1 examines the relationship between self-discrepant time al- location and work–family conflict. One concern related to the use of difference scores to assess discrepancy (see Edwards, 1994, for
detailed discussion) is the loss of information that occurs when computing a composite difference score from two variables. To address this concern, we followed the procedures recommended in Edwards (1995) for the use of difference scores as dependent variables, and also tested the relationships between WTF conflict and actual time allocation and preferred time allocation, sepa- rately. By reporting results separately for the composite self- discrepancy variable and its dimensions (actual and preferred time allocation), we addressed concerns related to loss of information (cf. Bono & Judge, 2003).
For H2 and H3, we employed polynomial regression and re- sponse surface methodology (Edwards & Parry, 1993) to test and illustrate the time allocation congruence and incongruence effects. This methodology unpacks perceptions of discrepancy into actual time allocations, preferred time allocations, and the difference between the two and is useful for examining phenomena such as person-job fit (see Kristof-Brown, Zimmerman, & Johnson, 2005, for review) in three dimensions.
In particular, we regressed actual time allocation (A), preferred time allocation (P), actual time allocation squared (A2), the inter- action between actual and preferred time allocations (A � P), and the preferred time allocation squared (P2) on each outcome vari- able (Z). The equation is:
Z � b0 � b1A � b2P � b3A 2 � b4(A � P) � b5P
2 � e.
For the polynomial regression analysis, both actual and pre- ferred time allocation variables were scaled by 100 to reduce the range for the higher order variables so that we could present meaningful coefficients within two decimal places while still ac- curately reflecting percentages (i.e., range 0 –1). Variables were centered to reduce collinearity in the higher order variables (Aiken & West, 1991).
Polynomial regression coefficients can be used to create a response surface that can be assessed along two lines of interest: congruence (P � A) and incongruence (P � –A). According to Edwards and Parry (1993), congruence effects exist when two conditions are met: (1) the higher order terms (A2, A � P, and P2) jointly and significantly explain variance in the dependent vari- able, and (2) the curvature along the line of incongruence is significantly different from zero. We tested H2 using these two criteria.
We used difference scores of self-discrepancy in our tests of mediation (Edwards & Lambert, 2007) as specified in H4 because it was unconventional to test for mediation using polynomial variables (see Colbert, Kristof-Brown, Bradley, & Barrick, 2008, for similar approach). Further, having first followed Edwards’s (1995) method for determining the significance of self-discrepant time allocation as it relates to WTF conflict, and then using polynomials to test relationships between time allocation and outcomes, we had insight into these relationships (i.e., whether actual time allocation, preferred time allocation, or the difference drives effects), and thus we used difference scores to test media- tion hypotheses.
Results
Table 2 shows the means, standard deviations, and intercorre- lations of the variables. WTF conflict is significantly correlated with self-discrepant time allocation for total, research, and service,
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
774 DAHM, GLOMB, MANCHESTER, AND LEROY
T ab
le 2
M ea
n s,
S ta
n d a rd
D ev
ia ti
o n s,
a n d
C o rr
el a ti
o n s
V ar
ia bl
e M
S D
1 2
3 4
5 6
7 8
9 10
11 12
13 14
15 16
17 18
19 20
21 22
23
1. A
ge 51
.1 5
11 .2
9 2.
A ss
is ta
nt pr
of es
so r
0. 28
0. 45
� .5
5 3.
A ss
oc ia
te pr
of es
so r
0. 30
0. 46
� .0
7 �
.4 1
4. H
ou rs
w or
ke d
54 .5
9 9.
83 �
.1 5
.0 3
� .0
1 5.
A dm
in is
tr at
iv e
du ti
es 0.
21 0.
40 .0
9 �
.1 6
.0 2
.0 7
6. P
ar tn
er em
pl oy
ed 0.
64 0.
48 �
.1 6
.0 5
.0 7
.0 0
.0 1
7. P
ar tn
er no
t em
pl oy
ed 0.
23 0.
42 .1
9 �
.0 9
� .1
2 �
.0 1
.0 3
� .7
1 8.
P ar
en t
0. 45
0. 50
� .3
8 .1
4 .0
8 .0
4 �
.0 1
.1 8
� .0
5 9.
F em
al e
0. 40
0. 49
� .1
7 .1
4 .1
2 �
.0 2
� .0
5 .0
8 �
.2 4
� .0
2 10
. W
or k-
to -f
am il
y co
nf li
ct 4.
77 1.
53 �
.2 1
.0 6
.0 7
.3 4
.1 0
.1 1
� .1
1 .1
3 .1
5 11
. T
ot al
ti m
e se
lf -d
is c.
28 .3
5 24
.7 5
� .0
7 .0
4 .0
9 .0
6 .1
1 �
.0 0
� .0
3 .0
2 .0
6 .2
2 12
. R
es ea
rc h
ti m
e se
lf -d
is cr
ep an
ta 9.
76 12
.9 7
� .0
6 �
.0 1
.0 7
.0 6
.0 6
� .0
0 �
.0 4
.0 3
.0 7
.1 6
.6 4
13 .
A ct
ua l
re se
ar ch
ti m
ea 36
.0 9
22 .2
6 �
.1 1
� .0
1 �
.1 1
.0 9
� .2
1 .0
6 .0
1 .0
6 �
.1 1
� .0
9 �
.3 8
� .4
0 14
. P
re fe
rr ed
re se
ar ch
ti m
ea 45
.8 5
20 .8
6 �
.1 5
� .0
2 �
.0 8
.1 4
� .1
8 .0
6 �
.0 1
.0 8
� .0
7 .0
0 �
.0 1
.2 0
.8 2
15 .
S er
vi ce
ti m
e se
lf -d
is cr
ep an
ta �
3. 34
8. 37
.0 7
.0 3
� .0
9 �
.0 4
.0 7
� .0
5 .0
4 �
.0 8
� .0
8 �
.1 0
� .2
6 �
.3 5
.1 5
� .0
6 16
. A
ct ua
l se
rv ic
e ti
m ea
17 .2
0 14
.2 9
� .0
5 .0
1 .0
7 .0
6 �
.0 4
.0 1
� .0
0 .0
9 .0
8 .1
5 .1
4 .1
3 �
.3 0
� .2
5 �
.5 9
17 .
P re
fe rr
ed se
rv ic
e ti
m ea
13 .8
6 11
.5 5
� .0
2 .0
4 .0
2 .0
5 .0
0 �
.0 3
.0 3
.0 6
.0 4
.1 2
� .0
2 �
.1 0
� .2
7 �
.3 5
� .0
0 .8
1 18
. T
ea ch
in g
ti m
e se
lf -d
is cr
ep an
ta �
4. 40
11 .6
5 .0
6 �
.1 0
� .0
3 �
.0 0
.1 8
.0 1
.0 5
� .0
5 �
.0 9
� .0
2 �
.3 1
� .5
4 .1
9 �
.1 3
� .1
5 .1
7 .1
1 19
. A
ct ua
l te
ac hi
ng ti
m ea
38 .7
4 21
.9 1
.0 4
.1 1
.0 6
� .1
4 �
.2 5
� .0
3 �
.0 6
� .0
7 .1
1 �
.0 5
.1 7
.2 4
� .5
8 �
.4 7
.1 3
� .2
6 �
.2 3
� .5
0 20
. P
re fe
rr ed
te ac
hi ng
ti m
ea 34
.3 4
18 .9
3 .0
8 .0
7 .0
4 �
.1 7
� .1
8 �
.0 4
� .0
5 �
.1 0
.0 7
� .0
7 .0
0 �
.0 6
� .5
5 �
.6 3
.0 6
� .2
0 �
.2 0
.0 3
.8 5
21 .
W or
k sa
ti sf
ac ti
on 0.
00 0.
69 .0
3 �
.0 1
� .0
9 �
.0 5
.0 2
� .0
1 .0
6 �
.0 2
� .0
2 �
.1 2
� .2
1 �
.1 4
.1 6
.0 8
.0 8
� .0
7 �
.0 2
.0 8
� .0
9 �
.0 5
22 .
P sy
ch ol
og ic
al w
el l-
be in
g 0.
00 0.
85 .1
8 �
.0 9
� .1
1 �
.1 8
.0 3
� .0
4 .1
1 �
.1 0
� .1
2 �
.4 4
� .2
4 �
.1 8
.1 1
.0 1
.0 7
� .0
7 �
.0 3
.1 0
� .0
7 �
.0 2
.4 0
23 .
P hy
si ca
l w
el l-
be in
g 3.
90 0.
91 .0
1 �
.0 3
� .0
9 �
.0 3
.0 5
.0 1
.0 2
� .0
4 �
.0 2
� .1
5 �
.1 5
� .0
9 .1
0 .0
6 .0
1 �
.0 1
� .0
1 .1
0 �
.1 0
� .0
6 .1
7 .4
0 24
. L
n( sa
la ry
) 11
.3 9
0. 41
.3 5
� .4
6 �
.1 8
.0 6
.1 3
� .0
9 .1
8 �
.1 1
� .2
2 �
.1 1
� .1
4 �
.0 9
.2 5
.2 1
.1 2
� .1
9 �
.1 5
.1 0
� .2
2 �
.1 9
.1 3
.1 5
.0 3
N o te
. N
s �
1, 06
9 to
1, 24
3. P
ai rw
is e
de le
ti on
us ed
fo r
m is
si ng
da ta
. C
or re
la ti
on s
si gn
if ic
an t
to p
� .0
5 ar
e it
al ic
iz ed
; p
� .0
1 ar
e un
de rl
in ed
an d
it al
ic iz
ed .
a V
ar ia
bl es
ar e
gi ve
n as
pe rc
en ta
ge s.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
775WORK–FAMILY CONFLICT AND TIME ALLOCATION
but not teaching. Participants’ preferred allocations suggested they would ideally spend the greatest amount of time researching (45.9%) as compared to service (13.9%) and teaching (34.3%). However, participants’ actual allocations suggested they allocate the greatest percentage of time to teaching (38.7%) as compared to research (36.1%) and service (17.2%). On average, faculty mem- bers would prefer to spend 9.8% more time researching, and 4.4% less time teaching and 3.3% less time doing service. Figures 2(a– c) depict scatter plots of individuals’ actual and preferred time allocations.
H1 suggests self-discrepant time allocation will increase with WTF conflict. As shown in Table 3, we find that WTF conflict relates to total self-discrepant time allocation (b � 3.51, p � .001). H1 further proposes WTF conflict is related to self-discrepant time allocation such that time allocated to work activities that require
higher self-regulatory resources (i.e., research) is less than pre- ferred, and time allocated to work activities that require fewer self-regulatory resources (i.e., teaching and service), is more than preferred. In support of H1, WTF conflict relates to self-discrepant research time allocation (b � 1.06, p � .001; Table 3) such that preferred is greater than actual research time. Also consistent with expectations, WTF conflict relates negatively to actual research time allocation (b � �1.69 p � .001), but WTF conflict does not significantly relate to preferred time allocation (b � �0.63, ns), suggesting the discrepancy is driven by the relationship with actual time allocation. Further supporting H1, WTF conflict relates neg- atively to self-discrepant service time allocation (b � �0.43, p � .024; Table 4), such that actual time to service is greater than preferred. It is interesting that WTF conflict relates positively to both actual service time (b � 1.46, p � .001), and to preferred service time (b � 1.03, p � .001) indicating that WTF conflict also may increase the desire to replenish resources through service time or may be a coping mechanism to bring actual and preferred time allocation more into alignment. Inconsistent with H1, WTF con- flict does not relate to actual or self-discrepant teaching time (see Table 5).
To further bolster our inferences, we examined our results with an alternative approach used to examine discrepancy (cf. Scott & Barnes, 2011). We regressed actual time allocations on WTF conflict controlling for preferred time allocations. WTF conflict relates to actual research time allocation (b � �1.15, p � .001) and actual service time allocation (b � 0.43, p � .026), but not actual teaching time allocation net of preferences, further indicat- ing that WTF conflict significantly relates to self-discrepant time allocation in the research and service domains.
H2 predicts that discrepancy between actual and preferred time allocation will be detrimental for work satisfaction, psychological well-being, and physical well-being. We find that total self- discrepancy significantly relates to work satisfaction (b � �0.01, p � .001), psychological well-being (b � �0.01, p � .001), and physical well-being (b � �0.01, p � .001). Table 6 presents the estimated coefficients for research time allocations as well as the slopes and curvatures along congruence and incongruence lines for the polynomial regressions predicting each dependent variable. For research, the addition of the three higher order variables signifi- cantly changes the R2 in the regression equation for each depen- dent variable (work satisfaction, psychological well-being, and physical well-being) suggesting that examining the response sur- faces is meaningful. For service, the addition of the three higher order variables significantly changes the R2 in the regression equation for psychological well-being, but not for work satisfac- tion or physical well-being (see Table 7). For teaching, the addi- tion of the three higher order variables significantly changes the R2
in the regression equations for psychological well-being and phys- ical well-being but not work satisfaction (see Table 8).
Figures 3(a– c; research), 4 (service), and 5 (a– b teaching) illustrate the response surfaces based on these coefficients. The lines of congruence and incongruence are shown on the bottom surface of the graph in Figure 3(a). The line of congruence extends from point (�.5, �.5) to point (.5, .5), while the line of incongru- ence extends from (�.5, .5) to (.5, �.5). For H2, we were primar- ily interested in the curvature along the line of incongruence because it showed how outcomes change as time allocations be- come more discrepant. For research time, the curvature along the
b
c
0
20
40
60
80
100
0 20 40 60 80 100
) %( e
miT ecivreS derrefer
P
Actual Service Time (%)
0
20
40
60
80
100
0 20 40 60 80 100
) %( e
miT g ni
hcaeT derrefer
P
Actual Teaching Time (%)
Actual Teaching Time = Preferred Teaching Time
0
20
40
60
80
100
0 20 40 60 80 100
) %( e
miT h craese
R derrefer
P
Actual Research Time (%)
Actual Research Time = Preferred Research Time
a
Actual Service Time = Preferred Service Time
Figure 2. Preferred versus actual (a) research time, (b) service time, and (c) teaching time. See the online article for the color version of this figure.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
776 DAHM, GLOMB, MANCHESTER, AND LEROY
line of incongruence was significant and in the expected direction for work satisfaction (p � .003), psychological well-being (p � .001), and physical well-being (p � .037) supporting H2(a– c). The curvature along the line of incongruence is negative, indicating an inverted-U (concave) shape suggesting outcomes decrease as ac- tual and preferred research time allocations become more discrep- ant. Similar patterns of results are observed for service and teach- ing. For service, the curvature along the line of incongruence is significant and negative for psychological well-being (p � .010). For teaching, the curvature along the line of incongruence is significant for psychological well-being (p � .001) and physical well-being (p � .001).
In line with Edwards and Cable (2009), we also examined the slope and intercept of each first principal axis (the ridge of the concave surfaces) to provide further support for the congruence effects. For concave surfaces, the first principal axis should run along the congruence line such that the slope is equal to 1 and the intercept is equal to 0 (Edwards & Cable, 2009). We used 10,000
bootstrap samples to construct 95% confidence intervals to test whether the 95% CI for slopes include 1 and the 95% CI for all intercepts include 0 for each response surface with significant curvature along the line of incongruence. We find that the 95% CI includes 1 for all slopes and 0 for all intercepts for 11 out of the 12 results tested. For the one exception, the 99% CI for the slope along the first principle axis for the relationship between research time and work satisfaction did include 1.
Overall, these results provide support for H2. When interpreting the response surfaces, it is important to exercise caution where there are few data points (see Figures 2a– c). In particular, for research, few individuals reported that actual time allocation far exceeds preferred, and for service and teaching, few individuals reported that preferred time allocation far exceeds actual and so those areas of the graphs should be interpreted with caution.
H3 proposes salary will increase with time allocation to activ- ities requiring self-regulatory resources. Examination of the time variables (A, P, A2, A � P and P2) reveals only the coefficient for
Table 4 Regressions of Work-to-Family Conflict on Self-Discrepant, Actual, and Preferred Service Time
Variables
Self-discrepant service time Actual service time Preferred service time
1 2 3 4 5 6
Constant �2.58 �1.34 4.50 0.29 1.92 �1.06 Age 0.04 0.03 0.06 0.09 0.10� 0.12��
Assistant professor 1.82� 1.71� 0.94 1.29 2.76� 3.00�
Associate professor �0.29 �0.30 2.39� 2.42� 2.10� 2.13�
Hours worked �0.04 �0.01 0.11� 0.04 0.08� 0.03 Administrative duties 1.87�� 2.00� �1.26 �1.69 0.61 0.30 Partner employed �0.81 �0.77 �0.30 �0.42 �1.11 �1.20 Partner not employed �0.22 �0.28 0.58 0.79 0.36 0.50 Parent �1.20� �1.11† 3.40�� 3.07�� 2.19�� 1.96�
Female �1.48� �1.27� 2.81�� 2.10� 1.33† 0.84 Work-to-family conflict �0.43� 1.46�� 1.03��
R2 0.03�� 0.04� 0.03�� 0.05�� 0.02� 0.04��
�R2 0.00� 0.02�� 0.02��
Note. Coefficients are unstandardized. † p � .10. � p � .05. �� p � .01.
Table 3 Regressions of Work-to-Family Conflict on Total Self-Discrepant Time and Self-Discrepant, Actual, and Preferred Research Time
Variables
Self-discrepant total time
Self-discrepant research time Actual research time Preferred research time
1 2 3 4 5 6 7 8
Constant 18.24� 8.10 7.48† 4.42 59.41�� 64.30�� 66.89�� 68.72��
Age 0.01 0.08 �0.04 �0.02 �0.51�� �0.54�� �0.54�� �0.55��
Assistant professor 4.12 4.96� �0.50 �0.25 �13.61�� �14.01�� �14.10�� �14.25��
Associate professor 5.34�� 5.42�� 0.68 0.71 �9.62�� �9.66�� �8.94�� �8.95��
Hours worked 0.10 �0.07 0.07 0.02 0.17� 0.25�� 0.24�� 0.27��
Administrative duties 7.25�� 6.20�� 1.38 1.07 �13.13�� �12.63�� �11.75�� �11.56��
Partner employed �1.83 �2.13 �1.50 �1.59 5.75�� 5.90�� 4.25� 4.31�
Partner not employed �1.78 �1.30 �1.60 �1.46 4.46† 4.23† 2.85 2.77 Parent 0.93 0.15 1.20 0.97 �0.67 �0.29 0.53 0.67 Female 1.75 0.06 1.29 0.78 �4.45�� �3.63� �3.16� �2.86�
Work-to-family conflict 3.51�� 1.06�� �1.69�� �0.63 R2 0.03�� 0.07�� 0.01 0.03�� 0.12�� 0.14�� 0.13�� 0.14��
�R2 0.04�� 0.01�� 0.01�� 0.00
Note. Coefficients are unstandardized. † p � .10. � p � .05. �� p � .01.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
777WORK–FAMILY CONFLICT AND TIME ALLOCATION
actual research time allocation (A) is significant (b � 0.35, p � .002), indicating that faculty salary is driven primarily by actual research time allocations regardless of allocation preferences or the discrepancy, supporting H3. Because the addition of the higher order variables significantly changes the R2 in the regression equation, we examine the relationship along the lines of incongru- ence and congruence. The maximum value along the line of incongruence is in the region where actual time allocation is greater than preferred (A � .84, P � .16), indicating that salary increases as actual time allocation increases up to this point.2
However, the slope along the line of incongruence is not signifi- cant (.41, p � .064). Examining the line of congruence, we find the slope is positive and significant (.28, p � .001) suggesting salary increases as actual and preferred research time increase jointly. As illustrated in Figure 6a, salary is much lower at the front left corner (�.5, �.5) of the figure, than at the back right corner (.5, .5) of the figure. The curvature along the line of congruence is also signif- icant (�1.16, p � .001), however, suggesting there may be dimin- ishing returns to allocating all of one’s time to research perhaps because it would preclude spending time on other work activities and be detrimental to salary.3,4
Though we did not hypothesize a relationship between salary and self-discrepant time allocation to activities requiring fewer self-regulatory resources, we explore it here. For service, the higher order polynomial variables are significant in predicting salary (see Table 7). The curvature along the line of incongruence is not significant, indicating that self-discrepant service time does not affect salary (see Figure 6b). The slope along the line of congruence is significant and in the opposite direction compared to research time, indicating salary decreases as preferred and actual service time increase jointly. Examination of the variables indi- cates salary is primarily affected by actual time allocated to service (b � �0.72, p � .007), rather than self-discrepancy.
The relationship between teaching time and salary is more complex (see Figure 6c). The addition of the higher order variables is significant (see Table 8). The curvature along the line of incon- gruence is negative and significant (p � .042), indicating that discrepancy from ideal time allocations to teaching corresponds
with lower salary, in contrast to research and service. Slope and curvature along the line of congruence are also negative and significant (ps � .001), suggesting salary decreases as actual and preferred time allocation to teaching increase jointly.
H4 proposes self-discrepancy mediates the relationship between WTF conflict and work satisfaction, psychological well-being, and physical well-being and H5 proposes actual time allocation to activities that require higher levels of self-regulatory resources mediates the relationship between WTF conflict and salary. We used path analysis (Edwards & Lambert, 2007; Preacher & Hayes, 2008) to test for mediation, describe the direct, indirect, and total effects of WTF conflict on outcomes, and test significance of the indirect effects by using 10,000 bootstrap samples to construct 95% confidence intervals, as suggested by Edwards and Lambert (2007). In support of H4a– c the indirect effects of WTF conflict through total self-discrepant time allocation on work satisfaction, psychological well-being, and physical well-being are significant (see Table 9); total self-discrepancy partially mediates the rela-
2 The maximum value was computed by taking the derivative of the equation with respect to A, setting it equal to 0, and solving for A (Edwards & Van Harrison, 1993). Because of the concave shape along the line of incongruence, this point is a maximum.
3 Regressions were completed with standard errors calculated at the individual level. We did not cluster the standard errors for the main analysis. Though individuals were nested in departments, we expected some heterogeneity within department and similarities that transcend de- partments to the university level making it unclear whether the department was the important factor. However, in supplemental analyses, we found the results were robust to clustering at the department level in that WTF conflict significantly relates to self-discrepant time allocation. For the relationship between time allocation and well-being outcomes, the results also were fairly robust to clustering at the department level such that the higher order variables are individually significant in predicting outcomes. Results available on request.
4 Though we operationalized time allocation in terms of percentages, when using an alternative measure in which we multiply percentages by total work hours, we found the relationships still supported (with the exception of the relationship between self-discrepant research time and physical well-being).
Table 5 Regressions of Work-to-Family Conflict on Self-Discrepant, Actual, and Preferred Teaching Time
Variables
Self-discrepant teaching time Actual teaching time Preferred teaching time
1 2 3 4 5 6
Constant �0.28 0.36 35.33�� 35.34�� 35.05�� 35.70��
Age �0.05 �0.06 0.35�� 0.35�� 0.30�� 0.30��
Assistant professor �3.17�� �3.23�� 11.85�� 11.85�� 8.68�� 8.62��
Associate professor �2.12� �2.13� 6.23�� 6.23�� 4.11�� 4.10��
Hours worked �0.03 �0.02 �0.23�� �0.23�� �0.26�� �0.25��
Administrative duties 4.43�� 4.50�� �12.44�� �12.44�� �8.00�� �7.94��
Partner employed 1.99† 2.01† �5.52� �5.52� �3.53† �3.51†
Partner not employed 1.77 1.74 �6.20� �6.20� �4.43� �4.46�
Parent �1.21 �1.16 �1.47 �1.47 �2.68� �2.63�
Female �1.11 �1.01 2.77† 2.77† 1.66 1.76 Work-to-family conflict �0.22 0.00 �0.22 R2 0.05�� 0.05�� 0.14�� 0.14�� 0.11�� 0.11��
�R2 0.00 0.00 0.00
Note. Coefficients are unstandardized. † p � .10. � p � .05. �� p � .01.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
778 DAHM, GLOMB, MANCHESTER, AND LEROY
tionship between WTF conflict and these outcomes.5 We con- ducted parallel analyses for self-discrepant time allocation in the research and service domains. As shown in Table 10, the indirect effect of WTF conflict through self-discrepant research time allo- cation is significant for work satisfaction (H4a) and psychological well-being (H4b) but not for physical well-being (H4c). We find that self-discrepant service time does not mediate the effects of WTF conflict on work satisfaction or well-being (see Table 11). Because WTF conflict did not significantly relate to self- discrepant teaching time, we did not test teaching time as a mediating mechanism.
In support of H5, the relationship between WTF conflict and salary is fully mediated by actual research time (b � �0.007, 95% CI [�.012, �.003]; see Table 10), such that the direct effects of WTF conflict on salary are no longer significant. Actual service time also mediates the relationship between WTF conflict and salary (b � �0.006, 95% CI [�.010, �.003]; see Table 11). Teaching time allocations were not significant mediators.
Gender results. H6 suggests women will be more likely than men to report self-discrepant time allocations as WTF conflict increases. As shown in Table 12, gender does not significantly moderate the relationship between WTF conflict and self- discrepant research, teaching, or total time. Gender does moderate the relationship between WTF conflict and self-discrepant service time allocation (b � �0.76, p � .04). Figure 7 shows the rela- tionship between WTF conflict and self-discrepant service time allocation is stronger for women such that they allocate more actual time to service than preferred as compared to men as WTF conflict increases.6 H6 is partially supported.
Longitudinal Analysis
To strengthen inferences from our primary analyses, we ana- lyzed data from the same population of employees approximately 2 years subsequent (Time 2) to our original survey (Time 1) and conducted analyses on the subset of our sample (N � 595) that participated in both surveys. This 2-year panel allowed us to (1) address concerns stemming from omitted variables related to the individual (Wooldridge, 2002) including the problem of single- source bias, (2) examine the lagged effects of WTF conflict on self-discrepant time allocation and self-discrepant time allocation on outcomes, and (3) assess the likelihood that our findings are driven by reverse causality (i.e., self-discrepant time allocation causes WTF conflict).
First, we conducted a first-difference regression to deal with omitted variable concerns (Wooldridge, 2002). This technique estimates our relationships of interest while removing (or differ- encing out) any effects driven by stable individual-level factors, such as personality. We found WTF conflict predicts total self- discrepant time (b � 2.28, p � .03) and self-discrepant research
5 We tested the sensitivity of the total self-discrepancy mediation effects to omitting each of the three domains in turn (research, teaching, and service). We found that total self-discrepant time allocation was a signif- icant mediator between WTF conflict and work satisfaction, psychological well-being, and physical well-being even if one of the three domains was omitted (i.e., total research and teaching, total research and service, or total service and teaching.)
6 WTF Conflict � Gender did not significantly relate to actual or preferred service time.
Table 6 Polynomial Regressions of Work Satisfaction, Well-Being, and Salary on Research Time Allocation
Variables
Work satisfaction Psychological
well-being Physical well-being Ln(salary)
1 2 3 4 5 6 7 8
Constant 0.29 0.31 0.46† 0.45 4.88�� 4.83�� 11.56�� 11.59��
Age 0.00 0.00 0.01� 0.01� �0.01� �0.01� 0.00 0.00 Assistant professor �0.05 �0.02 �0.01 �0.02 �0.28�� �0.33�� �0.51�� �0.47��
Associate professor �0.17�� �0.15�� �0.13† �0.12† �0.27�� �0.28�� �0.32�� �0.30��
Hours worked �0.01� �0.01� �0.02�� �0.02�� �0.01� �0.01† 0.00� 0.00 Admin duties 0.05 0.07 0.09 0.10 0.10 0.09 0.07�� 0.09��
Partner employed �0.01 �0.02 0.18� 0.17� 0.06 0.05 0.01 0.02 Partner not employed 0.02 0.02 0.24� 0.23� 0.06 0.05 0.07† 0.07�
Parent 0.03 0.03 �0.09 �0.09 �0.09 �0.08 �0.01 �0.02 Female 0.05 0.05 �0.11† �0.11† 0.04 0.04 �0.06�� �0.07��
Actual research time (A) 0.85�� 0.18 0.95�� 0.19 0.54� �0.02 0.39�� 0.35��
Preferred research time (P) �0.45� 0.07 �0.73�� 0.05 �0.22 0.45 0.03 �0.06 A2 �3.19�� �3.35�� �2.15� �0.55 A � P 3.99�� 6.20�� 4.58� �0.28 P2 �1.37 �2.58� �1.21 �0.33 R2 0.05�� 0.07�� 0.11�� 0.12�� 0.03�� 0.04�� 0.43�� 0.45��
�R2 0.02�� 0.01�� 0.01� 0.02��
Congruence line (P � A) Slope 0.25† 0.24 0.43� 0.28��
Curvature �.57 0.28 1.22� �1.16��
Incongruence line (P � �A) Slope 0.11 0.15 �0.48 0.41†
Curvature �8.55�� �12.12�� �7.94� �0.60
Note. Coefficients are unstandardized. �R2 refers to the change in explained variance attributable to the inclusion of the higher order variables: A2, A � P, and P2. † p � .10. � p � .05. �� p � .01.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
779WORK–FAMILY CONFLICT AND TIME ALLOCATION
time, b � 1.39, p � .011. These results provide confidence that our primary results are not driven by stable omitted third variables.
Second, we used a cross-lagged panel design to examine the lagged effects of our hypothesized model and address questions of reverse causality. To do this, we specified models for each out- come (work satisfaction, psychological well-being, and physical well-being) for each of the self-discrepant time allocation vari- ables. Each model included all variables at both Time 1 and Time 2 and was reciprocal in that it tested both hypothesized (i.e., WTF conflict predicts self-discrepancy; self-discrepancy predicts work satisfaction and well-being) and reverse causal relationships (i.e., self-discrepancy predicts WTF conflict; work satisfaction and well-being predict self-discrepancy; see Figure 8). All control variables from primary analyses were also included. Results avail- able on request.
We examined the lagged effects of Time 1 WTF on Time 2 self-discrepancy and Time 1 self-discrepancy on Time 2 outcomes, consistent with our theorized direction. Although WTF conflict at Time 1 significantly predicted total self-discrepant time allocation at Time 2, WTF conflict at Time 1 did not significantly predict the separate categories of self-discrepant time allocation to research, service, or teaching at Time 2.
Time 1 self-discrepant time allocations showed a mixed pattern of relationships with Time 2 outcomes. Time 1 total self-discrepant time allocation significantly predicted Time 2 physical well-being. Time 1 self-discrepant research time predicted Time 2 psycholog- ical and physical well-being. Time 1 self-discrepant service time predicted Time 2 psychological well-being. Time 1 self-discrepant teaching time did not relate to outcomes at Time 2. Time 1 actual
research time predicted Time 2 salary. In summary, there was some evidence for Time 1 variables to relate to Time 2 outcomes consistent with our theoretical propositions.
In addition to the lagged analyses that tested the relationships in a direction consistent with our hypotheses, we also examined the reverse causal lagged effects of Time 1 outcomes on Time 2 variables. We found no evidence that Time 1 self-discrepant time allocation, in total, or to research, service, or teaching, predicts Time 2 WTF conflict. With respect to the effects of Time 1 satisfaction and well-being outcomes on Time 2 self-discrepant time allocation, there were few significant effects suggesting re- verse causality. Time 1 work satisfaction and psychological well- being significantly predicted Time 2 self-discrepant service time and Time 1 physical well-being predicted Time 2 total self- discrepant time allocation. However, the relationship is in the opposite direction than expected; individuals reporting higher physical well-being at Time 1 reported more self-discrepant time allocation at Time 2. None of the Time 1 work satisfaction, psychological well-being, or physical well-being predicted Time 2 self-discrepant research or teaching time. Time 1 salary related to Time 2 actual service, but not research or teaching time. Collec- tively, our results do not provide compelling evidence for reverse causal relationships. However, it is important to note that our study was not optimally designed to examine causal direction.
Supplemental Analysis: Work Hours and Obligations
One alternative explanation for the relationship between WTF conflict and self-discrepant time allocation relates to differences in
Table 7 Polynomial Regressions of Work Satisfaction, Well-Being, and Salary on Service Time Allocation
Variables
Work satisfaction Psychological
well-being Physical well-being Ln(salary)
1 2 3 4 5 6 7 8
Constant 0.22 0.23 0.44 0.39 4.96�� 4.98�� 11.46�� 11.49��
Age 0.00 0.00 0.01� 0.01� �0.01�� �0.01�� 0.00 0.00 Assistant professor �0.10 �0.11 �0.04 �0.05 �0.33�� �0.35�� �0.57�� �0.55��
Associate professor �0.20�� �0.21�� �0.16� �0.17� �0.31�� �0.31�� �0.35�� �0.34��
Hours worked 0.00� 0.00� �0.02�� �0.02�� �0.01� �0.01� 0.00�� 0.00�
Administrative duties �0.01 �0.01 0.05 0.04 0.05 0.04 0.01 0.02 Partner employed 0.02 0.02 0.20� 0.20� 0.08 0.07 0.03 0.03 Partner not employed 0.05 0.05 0.26�� 0.26�� 0.08 0.07 0.09� 0.09�
Parent 0.03 0.03 �0.10† �0.10† �0.10 �0.10 0.00 �0.01 Female 0.04 0.04 �0.12� �0.11� 0.02 0.03 �0.07�� �0.08��
Actual service time (A) �0.48† 0.24 �0.30 1.21† �0.05 0.55 �0.45�� �0.72��
Preferred service time (P) 0.31 �0.61 0.31 �1.79† 0.16 0.11 0.04 0.02 A2 �1.89 �1.76 �1.29 �1.05†
A � P 5.35� 7.82� 4.39 0.52 P2 �3.85† �6.95�� �1.70 �0.39 R2 0.03�� 0.03�� 0.09�� 0.09�� 0.03�� 0.03�� 0.40�� 0.41��
�R2 0.00 0.01� 0.00 0.01��
Congruence line (P � A) Slope �0.37 �0.59 0.67 �0.69��
Curvature �0.39 �0.89 1.39 �0.93�
Incongruence Line (P � �A) Slope 0.85 3.00† 0.44 �0.74 Curvature �11.10� �16.53�� �7.38 �1.97
Note. Coefficients are unstandardized. �R2 refers to the change in explained variance attributable to the inclusion of the higher order variables: A2, A � P, and P2. † p � .10. � p � .05. �� p � .01.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
780 DAHM, GLOMB, MANCHESTER, AND LEROY
the total hours worked and the nature of obligations. If WTF conflict relates to working fewer hours in total (to meet family obligations), and research is considered more “discretionary” (Massy & Zemsky, 1994), individuals might allocate time to more pressing obligations during available work time (cf., Winslow, 2010). We addressed these concerns empirically and theoretically to rule out this alternative explanation.
First, we controlled for hours worked, such that effects are incremental to those related to work hours. In addition, we mea- sured time allocation in terms of percentages of time within the work domain; allocating less time to one activity means that more must be allocated to another activity within the work domain, not the family domain.
Second, we addressed the concern that WTF conflict relates to time allocation because of work obligations, rather than self- regulatory depletion. Research, teaching, and service are all obli- gations at the university from which we drew our sample making obligation alone an unlikely alternative explanation. However, the question of what obligations were more difficult to postpone may be relevant. Our supplemental survey indicated faculty believed that they were least able to postpone teaching requests, compared to postponing research and service (research: M � 3.26, SD � .90; service: M � 2.81, SD � .88; teaching: M � 2.37, SD � .69).7
Thus, if WTF conflict relates to time allocation because it pivots people to the most pressing obligation, then it should predict greater self-discrepancy with respect to teaching time. However, we found that WTF conflict did not significantly relate to actual, preferred, or self-discrepant teaching time.
Finally, we tested work hours as a mediating mechanism in line with Spector and Brannick (2011) who suggested, “the use of control variables would be far more productive, if approached as alternative hypothesis tests” (p. 297). If WTF conflict causes individuals to work fewer hours, which in turn causes individuals to allocate time to less discretionary tasks, WTF conflict should relate to fewer work hours, and work hours should mediate the relationship between WTF conflict and self-discrepant time allo- cations. We found that net of controls, WTF conflict significantly related to work hours (b � 2.05, p � .001), such that work hours actually increased with WTF conflict. Further, work hours were not significantly related to self-discrepant time allocation in total, or to research, service, or teaching; WTF conflict significantly related to self-discrepant time allocation whether or not we con- trolled for work hours. Given these threads of evidence, we con- cluded that work hours did not present a compelling alternative explanation for our results.
Supplemental Analysis: Academic Rank
Expectations and consequences for time allocation among re- search, service, and teaching were likely to vary according to academic rank. Accordingly, we conducted analyses by rank and found WTF conflict is associated with total self-discrepant time allocation for assistant, associate, and full professors. With respect
7 The survey item was “I can usually postpone [research, service, teach- ing] requests.”
Table 8 Polynomial Regressions of Work Satisfaction, Well-Being, and Salary on Teaching Time Allocation
Variables
Work satisfaction Psychological
well-being Physical well-being Ln(salary)
1 2 3 4 5 6 7 8
Constant 0.26 0.24 0.40 0.36 4.86�� 4.80�� 11.57�� 11.60��
Age 0.00 0.00 0.01� 0.01� �0.01� �0.01� 0.00 0.00 Assistant professor �0.06 �0.06 0.00 0.00 �0.29�� �0.29�� �0.54�� �0.50��
Associate professor �0.19�� �0.18�� �0.14� �0.12† �0.28�� �0.27�� �0.34�� �0.33��
Hours worked �0.01� �0.01� �0.02�� �0.02�� �0.01� �0.01† 0.00† 0.00 Administrative duties �0.05 �0.04 0.01 0.02 0.01 0.01 �0.01 0.00 Partner employed 0.00 0.01 0.18� 0.21� 0.06 0.09 0.02 0.02 Partner not employed 0.03 0.04 0.24� 0.27�� 0.06 0.08 0.07† 0.08�
Parent 0.02 0.01 �0.10† �0.12� �0.10 �0.11† �0.02 �0.03 Female 0.04 0.03 �0.12� �0.13� 0.03 0.02 �0.07�� �0.07��
Actual teaching time (A) �0.53�� �0.18 �0.55� 0.27 �0.53� 0.26 �0.16† �0.23†
Preferred teaching time (P) 0.30 0.03 0.37 �0.38 0.26 �0.36 �0.11 �0.28�
A2 �1.85� �3.09�� �2.74� �1.36��
A � P 4.03� 8.70�� 7.96�� 1.22 P2 �1.88 �5.16�� �4.32�� �1.10†
R2 0.03�� 0.04�� 0.09�� 0.11�� 0.03�� 0.05�� 0.40�� 0.42��
�R2 0.01 0.02�� 0.01�� .03��
Congruence line (P � A) Slope �0.15 �0.11 �0.10 �0.51��
Curvature 0.30 0.45 0.89† �1.24��
Incongruence line (P � �A) Slope �0.22 0.65 0.62 0.05 Curvature �7.76� �16.94�� �15.03�� �3.67�
Note. Coefficients are unstandardized. �R2 refers to the change in explained variance attributable to the inclusion of the higher order variables: A2, A � P, and P2. † p � .10. � p � .05. �� p � .01.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
781WORK–FAMILY CONFLICT AND TIME ALLOCATION
to allocation to specific activities, associate professors reported allocating less time to research than preferred in relation to WTF conflict (b � 1.56, p � .006), but the effects for assistant and full professors did not reach significance. This may be because assis- tant professors have the highest incentive to remain true to their research preferences given the dire consequences of not doing so. Full professors may have the most control over their time and ability to meet their preferences.
It is interesting that when we examined the role of rank in the relationship between self-discrepancy and outcomes, we found the relationship between self-discrepant research time allocation and work satisfaction was significant for full professors (�R2 � .02, p � .012) and associate professors (�R2 � .03, p � .018) but not for assistant professors (�R2 � .01, ns). In contrast, the effects of self-discrepant research time allocation on psychological well- being were strongest for assistant and associate professors (assis- tant: �R2 � .04, p � .010; associate: �R2 � .03, p � .013) and nonsignificant for full (full: �R2 � .01, ns). Because professors and associate professors are tenured, they may have higher expec- tations for achieving ideal time allocations, and deviance from ideals may cause more dissatisfaction with work. Because assistant professors will endure significant career consequences for not meeting time allocation preferences for research, deviation may cause worse psychological well-being.
Discussion
Academics in a research university setting know the rules of the game; the norms, expectations, and what it takes to succeed are typically shared and agreed on. And yet, among individuals expe- riencing high levels of WTF conflict (1 standard deviation above the mean), 73% reported allocating less actual time than preferred to their research endeavors. Organizations beyond academia have similar norms and expectations they rely on in their pursuit of performance (albeit not as neatly divided into research, teaching, service).
Many of us can relate to our findings that when WTF conflict is high, and our self-regulatory resources are low, we may find it difficult to deploy remaining resources to tasks that are high in complexity or those that delay gratification in support of our long-term goals. Instead, we gravitate toward activities that enable us to avoid further depletion or offer more immediate gratification; we may pivot to something less complex in which it is easier to establish a sense of closure or “quick win.” Our results suggest this self-discrepancy in time allocation has very real implications for work satisfaction, psychological well- being, physical well-being, and salary. Further, these time al- locations mediate the relationship between WTF conflict and work satisfaction, well-being and salary outcomes, making self-
Figure 3. Effects of actual and preferred (pref) research time (%) on (a) work satisfaction (sat), (b) psycho- logical (psych) well-being, and (c) physical (phys) well-being. The lines forming an “X” across the bottom of Figure 3a represent the lines of congruence and incongruence. The line of congruence extends from (–.5, –.5) to point (.5, .5). The line of incongruence extends from (–.5, .5) to point (.5, –.5). See the online article for the color version of this figure.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
782 DAHM, GLOMB, MANCHESTER, AND LEROY
discrepant time allocation a critical issue for both individual employees and organizations.
Theoretical and Empirical Contributions
In this paper we make important theoretical and empirical contributions to the work–family literature. First, the notion that WTF conflict may change how people allocate their time within their work roles is unique. Although measures of WTF conflict assess perceptions of time strains between work and family do- mains, and prior work shows that time investment in the work or family role reduces time devoted to the other role (e.g., Rothbard & Edwards, 2003), by examining time allocation within the work domain, we establish intrarole time allocations as an important piece of the work–family conflict puzzle. Although an intrarole approach to time allocation is uncharted in work–family research, it is not entirely new. Utilizing data from the 1999 National Study of Postsecondary Faculty, Winslow (2010) examined demographic and institutional predictors of time allocation at work. However,
this research did not examine relations between time allocations and outcomes such as well-being and salary, or work–family conflict as a driver of time allocations on the job.
Second, we provide a possible pathway for why negative work and life outcomes follow from WTF conflict, positing a work- related behavioral mediating mechanism. As noted by Cullen and Hammer (2007), “researchers need to develop a theoretical ratio- nale for how work-family conflict affects performance in order to examine those underlying mechanisms accordingly” (p. 270). We propose and find support for one potential reason why WTF conflict may be damaging to well-being and career success beyond the companion stressors of work and family demands. Feeling tapped out from WTF conflict may cause us to “take the easier road” and get something, anything, done to provide a sense of closure rather than expend the regulatory resources required to tackle the complex and longer term tasks necessary to achieve our goals.
Third, we contribute to the application of self-regulation theory to work–family issues. Although work–family issues have been framed using a resource perspective (cf. Allen, 2001; Grandey & Cropanzano, 1999) suggesting that resources (e.g., marital status, job tenure, family supportive work environment) can act as a determinant or buffer for WTF conflict, the influence of WTF conflict on self-regulatory resource depletion and consequences for work behaviors has not received extensive attention. Although our assessment is a retrospective global assessment of time allo- cation choices and has limitations (more on this later), it comple- ments existing work that manipulates choice and depletion in short-term laboratory settings by examining presumably more chronic, as compared to state, depletion. Moreover, the activities studied in the lab are often not relevant to work settings (e.g., holding one’s arm in ice water, Vohs et al., 2008) or require a choice between a limited set of activities categorized as work and play (e.g., persist on a task or play video games, Vohs et al., 2008) rather than choices among different types of work activities. It is this latter scenario that is most relevant in organizational settings.
Gender Effects
Our interaction results suggest that as WTF conflict increases, women are more likely than men to become self-discrepant in their
Figure 4. Effects of actual and preferred (pref) service time (%) on psychological (psych) well-being. See the online article for the color version of this figure.
Figure 5. Effects of actual and preferred (pref) teaching time (%) on (a) psychological (psych) well-being and (b) physical (phys) well-being. See the online article for the color version of this figure.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
783WORK–FAMILY CONFLICT AND TIME ALLOCATION
time allocations. WTF conflict may be more depleting for women because allowing work to interfere with family is a violation of social role expectations for women, but not for men; strong role pressures have more sanctions for noncompliance (Greenhaus & Powell, 2003; Kahn et al., 1964). Women may become more self-discrepant in service, specifically, when WTF conflict is high because they may find it to be especially replenishing. Our analysis of faculty perceptions found that, as compared to men, women report service activities provide a greater sense of closure (Mdifference � 1.39, p � .001, SEdifference � .38) and are more prosocial (Mdifference � 0.80, p � .008, SEdifference � .28). Further women reported feeling significantly worse than men (Mdifference � 0.79, p � .01, SEdifference � .28) saying “no” to service requests,
indicating that they may find it more difficult to refuse service requests from others when WTF conflict is high.
Our WTF conflict and gender interaction findings should be considered in light of the direct relationships between gender and actual and preferred time. We see in Table 4 that gender signifi- cantly relates to self-discrepant service time; this effect is primar- ily driven by differences in actual service time, rather than pre- ferred, suggesting women may be more likely to comply with service requests despite preferences. This is consistent with qual- itative work by Misra and colleagues (Misra, Lundquist, Holmes, & Agiomavritis, 2011) suggesting women in academia are partic- ularly likely to be burdened by service obligations. Our results also are consistent with the COACHE (2010) Tenure-Track Faculty Job
Table 9 Path-Analytic Results: Indirect and Total Effects of Work-to-Family Conflict Via Total Self-Discrepant Time Allocation on Work Satisfaction and Well-Being
Variables PMX PYM Direct effects
(PYX) Indirect effects (PYM � PMX)
Total effects (PYX � (PYM � PMX))
Simple paths for work satisfaction 3.515�� �.005�� �.028† �.018� �.046��
Simple paths for psychological well-being 3.515�� �.005�� �.197�� �.018� �.214��
Simple paths for physical well-being 3.524�� �.004�� �.083�� �.015� �.098��
Note. PMX � path from X (work–to–family conflict) to M (total self– discrepant time allocation); PYM � path from M (total self– discrepant time allocation) to Y (work satisfaction and well-being); PYX � path from X to Y (e.g., the direct effect of work-to-family conflict on work satisfaction and well-being); PYM � PMX � indirect effects; PYX � (PYM � PMX) � total effect of X on Y. † p � .10. � p � .05. �� p � .01.
Figure 6. Effects of actual and preferred (pref): (a) research time (%), (b) service time (%), and (c) teaching time (%) on ln(salary). See the online article for the color version of this figure.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
784 DAHM, GLOMB, MANCHESTER, AND LEROY
Satisfaction Survey indicating women are less satisfied than men with the way they spend their time as faculty members and the amount of time to conduct research. More important, in our results, there were no gender differences in total work hours so it seems that the interplay of time allocation and gender is important.
These gender effects are especially important to consider given the persistent gender gap in salary. The gender gap in salary has diminished, but net of human capital, remains largely unexplained (Blau & Kahn, 2006). Gender discrimination provides one expla- nation for the remaining gap, but we provide a potential behavioral explanation as well. If women allocate less time than men to tasks requiring more self-regulatory resources, in this case preferring and actually allocating less time to research (Table 3), and they are more easily nudged from these preferences when experiencing WTF conflict, this could help explain pay disparity because addi- tional time spent on research translates into research performance and, ultimately, career rewards.
Examination of Alternative Mechanisms
To provide additional support for our findings, we examined a number of alternative explanations to deepen our insights. A compel- ling alternative explanation relates to total hours worked; if an em- ployee works fewer hours in total, employees might necessarily allocate more time to less discretionary requirements. Thus, it is important that our analyses controlled for hours worked. In addition, we found no support for work hours as a mediator of the WTF conflict and time allocation relationship, suggesting this is not a likely alter-
native and bolstering propositions stemming from self-discrepancy and self-regulation theories. Future studies could examine whether WTF conflict relates to self-discrepancy between preferred and actual total work hours.
We also used data from two time points to strengthen inferences from our primary analyses. Our first difference regression analysis examines within individual changes and shows changes in WTF conflict relate to changes in self-discrepant time allocation, making omitted variable explanations stemming from personal characteristics unlikely. Further, our cross-lagged panel design analyses showed that self-discrepant time allocation at Time 1 has lagged effects on out- comes at Time 2, but found little support for reverse causal explana- tions. These lagged results suggest the consequences of self- discrepant time allocation may persist; additional attention to the duration of effects would be intriguing.
Implications
Understanding how individuals allocate time when work–family conflict is high and self-regulatory resources are depleted may help employees and managers structure jobs and work time more effec- tively. Organizations would do well to manage the flow of work tasks based on the level of required self-regulatory resources, or provide ways to break complex, longer term tasks into less complex tasks that may be completed for more immediate gratification; research has shown the effects of goal setting on performance are greater for simple as compared to complex tasks (Wood, Mento, & Locke, 1987). In this way, employees will be more likely to both enter into and
Table 10 Path-Analytic Results: Indirect and Total Effects of Work-to-Family Conflict Via Research Time Allocation on Work Satisfaction, Well-Being, and Salary
Variables PMX PYM Direct effects
(PYX) Indirect effects (PYM � PMX)
Total effects (PYX � (PYM � PMX))
Simple paths for work satisfaction 1.061�� �.007�� �.039� �.007� �.046��
Simple paths for psychological well-being 1.061�� �.006�� �.208�� �.007� �.214��
Simple paths for physical well-being 1.064�� �.003 �.094�� �.003 �.098��
Simple paths for salary �1.831�� .004�� �.009 �.007� �.016�
Note. Simple paths for work satisfaction and well-being are through self-discrepant research time, whereas simple paths for salary are through actual research time. PMX � path from X (work-to-family conflict) to M (research time allocation mediators); PYM � path from M (research time allocation mediators) to Y (work satisfaction, well-being, and salary); PYX � path from X to Y (e.g., the direct effect of work-to-family conflict on work satisfaction, well-being, and salary); PYM � PMX � indirect effects; PYX � (PYM � PMX) � total effect of X on Y. � p � .05. �� p � .01.
Table 11 Path-Analytic Results: Indirect and Total Effects of Work-to-Family Conflict Via Service Time Allocation on Work Satisfaction, Well- Being, and Salary
Variables PMX PYM Direct effects
(PYX) Indirect effects (PYM � PMX)
Total effects (PYX � (PYM � PMX))
Simple paths for work satisfaction �0.430� .004† �.044�� �.002 �.046��
Simple paths for psychological well-being �0.430� .001 �.214�� �.000 �.214��
Simple paths for physical well-being �0.429� �.001 �.098�� .000 �.098��
Simple paths for salary 1.502�� �.004�� �.010 �.006� �.016�
Note. Simple paths for work sat and well– being are through self– discrepant service time, while simple paths for salary are through actual service time. PMX � path from X (work-to-family conflict) to M (research time allocation mediators); PYM � path from M (research time allocation mediators) to Y (work satisfaction, well– being, and salary); PYX � path from X to Y (e.g., the direct effect of work-to-family conflict on work satisfaction, well-being, and salary); PYM � PMX � indirect effects; PYX � (PYM � PMX) � total effect of X on Y. † p � .10. � p � .05. �� p � .01.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
785WORK–FAMILY CONFLICT AND TIME ALLOCATION
remain engaged in tasks that may otherwise be avoided due to required self-regulatory resources; this may serve to better align actual time allocation with individual and organization preferences, making individuals more satisfied and increasing the likelihood of career rewards. This implication is in line with the “power of small wins” research showing that making progress on meaningful work is key to motivation (Amabile & Kramer, 2011). From an individual perspec- tive, highlighting the effects of work–family conflict depletion on time allocation to work tasks will empower individuals to recognize and correct these allocations to be more congruent with preferences. Individuals with high job control can protect and rebuild self- regulatory resources by breaking their own work into more short- term, discrete tasks. This idea is consistent with research that indicates that breaking tasks into 90-min intervals is an effective way to be productive (Ericsson, Krampe, & Tesch-Romer, 1993).
Our results also may suggest promising means of work recovery at work. Research on work recovery has primarily focused on work recovery outside of work time or by taking breaks within work time, but there may be opportunities for work recovery through the work
activities themselves. Indeed, work recovery researchers have made a call for a better understanding of how work recovery might be accomplished at work (Trougakos, Hideg, Cheng, & Beal, 2013). From a managerial perspective, activities that rebuild regulatory re- sources, or at the very least do not further tap resources, could be assigned and appropriately interspersed among resource-depleting tasks, or could be bundled with resource-depleting tasks (e.g., Milk- man, Minson, & Volpp, 2013). Individuals experiencing high levels of WTF conflict also may wish to seek replenishment through non- work activities that provide psychological detachment from work, relaxation (Sonnentag & Fritz, 2007), or autonomous social interac- tions (Trougakos et al., 2013) so that they are better able to allocate time in accordance with ideals at work, as variables related to self-care such as lack of sleep have been associated with ego depletion (Barnes, Schaubroeck, Huth, & Ghumman, 2011). Dynamics related to time also must be considered, as something that is restorative in the short term may have different outcomes over the long term.
Organizations also might encourage the careful planning and pro- tection of time allocations, especially for employees who may be especially vulnerable to self-discrepant time allocations. Employees and organizations might consider blocking time for activities that require greater self-regulatory resources during the part of the day (or week) when they are less depleted (such as the morning). Organiza- tions could facilitate this by scheduling meetings or requests on employee time when depletion is likely highest. The strategic use of time and work management tools such as blocking time for complex tasks or “parking downhill” (i.e., identifying small tasks that you will start the next day; Bolker, 1998) each day may be effective.
Limitations and Future Directions
One potential limitation is the possibility of reverse causality. Our primary analyses were conducted on cross-sectional data and although our analyses across two time points do not fully answer the issue of causality, they do strengthen our inferences. This study was not designed to test causal direction, but in our examination of reverse causality using a cross-lagged panel design we found
Table 12 Regressions of Work-to-Family Conflict � Gender on Self-Discrepant Time
Variables
Self-discrepant total time
Self-discrepant research time
Self-discrepant service time
Self-discrepant teaching time
1 2 3 4 5 6 7 8
Constant 8.10 7.61 4.42 6.17 �1.34 �2.75 0.36 �0.08 Age 0.08 0.08 �0.02 �0.02 0.03 0.04 �0.06 �0.06 Assistant professor 4.96� 4.95� �0.25 �0.23 1.71� 1.70† �3.23�� �3.23��
Associate professor 5.42�� 5.44�� 0.71 0.65 �0.30 �0.25 �2.13� �2.11�
Hours worked �0.07 �0.07 0.02 0.02 �0.01 �0.02 �0.02 �0.02 Administrative duties 6.20�� 6.19�� 1.07 1.10 2.00�� 1.96�� 4.50�� 4.49��
Partner employed �2.13 �2.11 �1.59 �1.66 �0.77 �0.72 2.01† 2.03†
Partner not employed �1.30 �1.26 �1.46 �1.60 �0.28 �0.17 1.74 1.77 Parent 0.15 0.13 0.97 1.01 �1.11† �1.14† �1.16 �1.17 Female 0.06 1.38 0.78 �3.87 �1.27� 2.49 �1.01 0.17 Work-to-family conflict 3.51�� 3.61�� 1.06�� 0.73� �0.43� �0.16 �0.22 �0.14 Work-to-Family Conflict � Gender �0.27 0.94† �0.76� �0.24 R2 0.07�� 0.07�� 0.03�� 0.03�� 0.04�� 0.04�� 0.05�� 0.05��
�R2 0.00 0.00† 0.00� 0.00
Note. Coefficients are unstandardized. † p � .10. � p � .05. �� p � .01.
-6
-5
-4
-3
-2
-1
0 WTF low WTF high
Se lf
-d is
cr ep
an t S
er vi
ce T
im e
Men
Women
Figure 7. Self-Discrepant Service Time � Gender Interaction. WTF � work-to-family conflict.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
786 DAHM, GLOMB, MANCHESTER, AND LEROY
evidence for the hypothesized direction and little evidence of the reverse relationship.
Although we did not find empirical evidence of the reverse relationship, we can examine the issue theoretically and consider whether self-discrepant time allocation might increase perceptions of WTF conflict. Thompson and Bunderson (2001) suggested, but did not test, the idea that allocating time out of alignment with preferences at work decreases resources available for the family domain and increases perceptions of WTF conflict. Experiencing persistent self-discrepant time allocation at work over an extended time period may cause an individual to look for reasons to ratio- nalize this uncomfortable self-discrepancy to create a sense of consistency in the self-concept (Weick, 1995). Self-discrepant time allocation at work may cause an increase in perceptions that family is interfering with work, to provide an explanation for falling short of work goals. However, we examined the relation- ship with WTF (rather than family-to-work) conflict, making this reverse causal explanation less plausible. Finally, reverse causal explanations may be more likely in an environment in which employees have less control over their time and time allocation decisions are made by others. The faculty members in our sample have a relatively high degree of job control, thus our examination of the internal factors that would lead to self-discrepant time allocation seems more relevant, rather than the reverse.
However, we acknowledge that this is complex and the rela- tionship may be reciprocal, consistent with literature suggesting that depletion may lead to loss spirals (i.e., conservation of re- sources theory, Hobfoll, 1989). Future work might examine how changes in work or family demands (e.g., birth of a child, promo- tion) influence self-discrepant time allocations, or examine these relationships in a more episodic nature, to provide insight into the causal direction.
Our findings may not generalize to all professions. Faculty members at research institutions have a high degree of job control and control over how they allocate their time. Most professionals have some discretion over their time allocation, in part due to recent changes in technology that provide greater flexibility in terms of when, where, and how work gets done (Kossek, Lautsch,
& Eaton, 2006). Employees who have less control, such as factory workers or call center representatives, may not have the latitude to adjust time allocations in response to depletion. Thus, control may be a key issue and may explain why WTF conflict did not signif- icantly relate to teaching time; time spent in the classroom is less discretionary compared to research and service responsibilities. Also, the academic institution where we conducted our study rewards tasks that require greater levels of self-regulatory re- sources (i.e., research). Had we tested our hypotheses in a sample where tasks with immediate gratification and sense of closure are more valued, perhaps customer service or nursing, our findings may be different.
Our study limitations also included measurement limitations. In this study, we focused on WTF conflict, rather than family-to- work conflict for theoretical and empirical reasons. In a supple- mental analysis, we found that family-to-work conflict (two items from Netemeyer et al., 1996; r � .73) had a smaller and less significant relationship with total self-discrepant time allocation (b � 1.10, p � .025) as compared to WTF conflict’s relationship with total self-discrepant time allocation (b � 3.51, p � .01). Our study is also limited by single item measures. Although single item measures can be valid for global perceptions of constructs such as we have here (Wanous, Reichers, & Hudy, 1997) and single item measures of health (DeSalvo et al., 2006) have been shown to be valid in prior research, lengthier measures would be ideal.
Our final limitation relates to our measurement of depletion of self-regulatory resources. We conceptualize WTF conflict as de- pleting, and self-discrepant time allocation as a behavioral mani- festation of that depletion. In many ways, this conceptualization is consistent with popular depletion paradigms in laboratory studies, which impose depleting manipulations (e.g., eating radishes in- stead of tempting chocolates), and then employ behavioral, rather than affective, measures to infer levels of depletion (e.g., persis- tence on subsequent unsolvable geometric puzzles; Baumeister et al., 1998). We conducted a parallel, but more naturalistic study, using WTF conflict as a depleting mechanism, and time allocation behaviors at work to measure the results of that depletion. None- theless, research in which depletion is captured more directly is
Time 1 Time 2
Time Allocation
Psych Well-being Work Satisfaction
Physical Well-being Salary
Psych Well-being Work Satisfaction
Physical Well-being Salary
Work-to-family Conflict
Work-to-family Conflict
Time Allocation
Hypothesized Direction
Reverse Causal
Figure 8. Cross-lagged panel models. For well-being and work satisfaction models, time allocation � self-discrepant time; for the salary model, time allocation � actual time; Psych � psychological.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
787WORK–FAMILY CONFLICT AND TIME ALLOCATION
needed. An episodic approach could detect episodes of WTF conflict and concurrent depletion and link that to time allocation to work activities using experience sampling. Our conceptualization of WTF conflict was as a “level” of conflict, and research has called for increased attention to “episodes” of conflict (Maertz & Boyar, 2011). Such a design would allow for a more powerful test of our key hypotheses and bolster propositions about causal direc- tion.
In conclusion, we find WTF conflict relates to time allocation at work, which mediates effects on important well-being and career success outcomes. By highlighting WTF conflict’s relationship to time allocation at work, we give individuals and organizations actionable insight into a mechanism through which WTF conflict may take its toll. Forewarned is forearmed, and employees and organizations might be well poised to regularly ask themselves: Is that how you want to spend your time? As the saying goes, “Time is the coin of your life. It is the only coin you have, and only you can determine how it will be spent” (Sandburg, 2015).
References
Aarts, H. (2007). On the emergence of human goal pursuit: The noncon- scious regulation and motivation of goals. Social and Personality Psy- chology Compass, 1, 183–201. http://dx.doi.org/10.1111/j.1751-9004 .2007.00014.x
Agarwal, N. C. (1981). Determinants of executive education. Industrial Relations, 20, 36 – 46.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage.
Albert, S. (1977). Temporal comparison theory. Psychological Review, 84, 485–503. http://dx.doi.org/10.1037/0033-295X.84.6.485
Allen, T. D. (2001). Family-supportive work environments: The role of organization perceptions. Journal of Vocational Behavior, 58, 414 – 435. http://dx.doi.org/10.1006/jvbe.2000.1774
Allen, T. D., Herst, D. E. L., Bruck, C. S., & Sutton, M. (2000). Conse- quences associated with work-to-family conflict: A review and agenda for future research. Journal of Occupational Health Psychology, 5, 278 –308. http://dx.doi.org/10.1037/1076-8998.5.2.278
Amabile, T. M., & Kramer, S. J. (2011, May). The power of small wins. Harvard Business Review, 70 – 80.
Amstad, F. T., Meier, L. L., Fasel, U., Elfering, A., & Semmer, N. K. (2011). A meta-analysis of work-family conflict and various outcomes with a special emphasis on cross-domain versus matching-domain rela- tions. Journal of Occupational Health Psychology, 16, 151–169. http:// dx.doi.org/10.1037/a0022170
Bargh, J. A. (1990). Auto-motives: Preconscious determinants of social interaction. In E. T. Higgins & R. M. Sorrentino (Eds.), Handbook of motivation and cognition: Vol. 2: Foundations of social behavior. New York, NY: Guilford Press.
Barnes, C. M., Schaubroeck, J., Huth, M., Ghumman, S. (2011). Lack of sleep and unethical conduct. Organizational Behavior and Human De- cision Processes, 115, 169 –180.
Baumeister, R. F. (2002). Yielding to temptation: Self-control failure, impulsive purchasing, and consumer behavior. The Journal of Consumer Research, 28, 670 – 676. http://dx.doi.org/10.1086/338209
Baumeister, R. F., Bratslavsky, E., Muraven, M., & Tice, D. M. (1998). Ego depletion: Is the active self a limited resource? Journal of Person- ality and Social Psychology, 74, 1252–1265. http://dx.doi.org/10.1037/ 0022-3514.74.5.1252
Baumeister, R. F., & Heatherton, T. F. (1996). Self-regulation failure: An overview. Psychological Inquiry, 7, 1–15. http://dx.doi.org/10.1207/ s15327965pli0701_1
Baumeister, R. F., Heatherton, T. F., & Tice, D. M. (1994). Losing control: How and why people fail at self-regulation. San Diego, CA: Academic.
Baumeister, R. F., & Vohs, K. D. (2007). Self-regulation, ego depletion, and motivation. Social and Personality Psychology Compass, 1, 115– 128. http://dx.doi.org/10.1111/j.1751-9004.2007.00001.x
Bellas, M. L., & Toutkoushian, R. K. (1999). Faculty time allocations and research productivity: Gender, race and family effects. The Review of Higher Education, 22, 367–390. http://dx.doi.org/10.1353/rhe.1999 .0014
Bishop, J. (1987). The recognition and reward of employee performance. Journal of Labor Economics, 5(Suppl. 4), S36 –S56. http://dx.doi.org/ 10.1086/298164
Blau, F. D., & Kahn, L. M. (2006). The gender pay gap: Going, going, . . . but not gone. In F. Blau, M. Brinton, & D. Grusky (Eds.), The declining significance of gender? (pp. 37– 66). New York, NY: Russell Sage Foundation.
Bolino, M. C., & Turnley, W. H. (2005). The personal costs of citizenship behavior: The relationship between individual initiative and role over- load, job stress, and work-family conflict. Journal of Applied Psychol- ogy, 90, 740 –748.
Bolker, J. (1998). Writing your dissertation in fifteen minutes a day. New York, NY: Holt.
Bono, J. E., Glomb, T. M., Shen, W., Kim, E., & Koch, A. J. (2013). Building positive resources: Effects of positive events and positive reflection on work-stress and health. Academy of Management Journal, 56, 1601–1627. http://dx.doi.org/10.5465/amj.2011.0272
Bono, J., & Judge, T. A. (2003). Self-concordance at work: Toward understanding the motivational effects of transformational leaders. Academy of Management Journal, 46, 554 –571. http://dx.doi.org/ 10.2307/30040649
Bound, J., & Johnson, G. (1992). Changes in the structure of wages in the 1980’s: An evaluation of alternative explanations. The American Eco- nomic Review, 82, 371–392.
Brunstein, J. C. (1993). Personal goals and subjective well-being: A longitudinal study. Journal of Personality and Social Psychology, 65, 1061–1070.
Bureau of Labor Statistics. (2010). American Time Use Survey. www.bls .gov/tus
Campbell, D. J. (1988). Task complexity: A review and analysis. Academy of Management Review, 13, 40 –52.
Carlson, D. S., Kacmar, K. M., & Williams, L. J. (2000). Construction and initial validation of a multidimensional measure of work–family con- flict. Journal of Vocational Behavior, 56, 249 –276. http://dx.doi.org/ 10.1006/jvbe.1999.1713
Carl Sandburg, C. (2015, January). BrainyQuote.com. Retrieved from website: http://www.brainyquote.com/quotes/quotes/c/carlsandbu121791.html
Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positive and negative affect: A control-process view. Psychological Review, 97, 19 –35.
Chartrand, T. L., & Bargh, J. A. (1996). Automatic activation and impres- sion formation and memorization goals: Nonconscious goal priming reproduces effects of explicit task instructions. Journal of Personality and Social Psychology, 71, 464 – 478. http://dx.doi.org/10.1037/0022- 3514.71.3.464
Cinamon, R. G., & Rich, Y. (2002). Profiles of attribution of importance to life roles and their implications for the work-family conflict. Journal of Counseling Psychology, 49, 212–220.
COACHE. (2010). Tenure-track faculty job satisfaction survey. Cam- bridge, MA: Harvard University Graduate School of Education. Re- trieved December 20, 2013, from http://isites.harvard.edu/fs/docs/icb .topic1023643.files/COACHE_AnalysisByAcademicAreaAndGender_ Report.pdf
Colbert, A. E., Kristof-Brown, A., Bradley, B. H., & Barrick, M. R. (2008). CEO transformational leadership: The role of goal importance congru-
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
788 DAHM, GLOMB, MANCHESTER, AND LEROY
ence in top management teams. Academy of Management Journal, 51, 81–96. http://dx.doi.org/10.5465/AMJ.2008.30717744
Cullen, J. C., & Hammer, L. B. (2007). Developing and testing a theoret- ical model linking work-family conflict to employee safety. Journal of Occupational Health Psychology, 12, 266 –278. http://dx.doi.org/ 10.1037/1076-8998.12.3.266
Demerouti, E., Bakker, A. B., & Bulters, A. J. (2004). The loss spiral of work pressure, work-home interface and exhaustion: Reciprocal rela- tions in a three-wave study. Journal of Vocational Behavior, 64, 131– 149. http://dx.doi.org/10.1016/S0001-8791(03)00030-7
Demerouti, E., Bakker, A. B., & Schaufeli, W. B. (2005). Spillover and crossover of exhaustion and life satisfaction among dual-earner parents. Journal of Vocational Behavior, 67, 266 –289. http://dx.doi.org/10.1016/ j.jvb.2004.07.001
DeSalvo, K. B., Fisher, W. P., Tran, K., Bloser, N., Merrill, W., & Peabody, J. (2006). Assessing measurement properties of two single- item general health measures. Quality of Life Research: An International Journal of Quality of Life Aspects of Treatment, Care and Rehabilita- tion, 15, 191–201. http://dx.doi.org/10.1007/s11136-005-0887-2
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49, 71– 75. http://dx.doi.org/10.1207/s15327752jpa4901_13
Duxbury, L. E., & Higgins, C. A. (1991). Gender differences in work- family conflict. Journal of Applied Psychology, 76, 60 –74. http://dx.doi .org/10.1037/0021-9010.76.1.60
Eagly, A. H. (1987). Sex differences in social behavior: A social-role interpretation. Hillsdale, NJ: Erlbaum.
Edwards, J. R. (1994). The study of congruence in organizational behavior research: Critique and proposed alternative. Organizational Behavior and Human Decision Processes, 58, 51–100. http://dx.doi.org/10.1006/ obhd.1994.1029
Edwards, J. R. (1995). Alternatives to difference scores as dependent variables in the study of congruence in organizational research. Orga- nizational Behavior and Human Decision Processes, 64, 307–324. http://dx.doi.org/10.1006/obhd.1995.1108
Edwards, J. R., & Cable, D. M. (2009). The value of value congruence. Journal of Applied Psychology, 94, 654 – 677.
Edwards, J. R., & Lambert, L. S. L. (2007). Methods for integrating moderation and mediation: A general analytical framework using mod- erated path analysis. Psychological Methods, 12, 1–22. http://dx.doi.org/ 10.1037/1082-989X.12.1.1
Edwards, J. R., & Parry, M. E. (1993). On the use of polynomial regression equations as an alternative to difference scores in organizational re- search. Academy of Management Journal, 36, 1577–1613. http://dx.doi .org/10.2307/256822
Edwards, J. R., & Van Harrison, R. (1993). Job demands and worker health: Three-dimensional reexamination of the relationship between person-environment fit and strain. Journal of Applied Psychology, 78, 628 – 648.
Elliot, A. J., & Sheldon, K. M. (1997). Avoidance achievement motivation: A personal goals analysis. Journal of Personality and Social Psychol- ogy, 3, 171–185.
Elliot, A. J., Sheldon, K. M., & Church, M. (1997). Avoidance personal goals and subjective well-being. Personality and Social Psychology Bulletin, 23, 915–927.
Ericsson, K., Krampe, R., & Tesch-Romer, C. (1993). The role of delib- erate practice in the acquisition of expert performance. Psychological Review, 100, 363– 406. http://dx.doi.org/10.1037/0033-295X.100.3.363
Frone, M. R., Russell, M., & Cooper, M. L. (1992). Antecedents and outcomes of work-family conflict: Testing a model of the work-family interface. Journal of Applied Psychology, 77, 65–78. http://dx.doi.org/ 10.1037/0021-9010.77.1.65
Gable, S. L., Reis, H. T., Impett, E. A., & Asher, E. R. (2004). What do you do when things go right? The intrapersonal and interpersonal benefits of
sharing positive events. Journal of Personality and Social Psychology, 87, 228 –245. http://dx.doi.org/10.1037/0022-3514.87.2.228
Gailliot, M. T., & Baumeister, R. F. (2007). Self-regulation and sexual restraint: Dispositionally and temporarily poor self-regulatory abilities contribute to failures at restraining sexual behavior. Personality and Social Psychology Bulletin, 33, 173–186. http://dx.doi.org/10.1177/ 0146167206293472
Galinsky, E., Kim, S. S., & Bond, J. T. (2001). Feeling overworked: When work becomes too much. New York, NY: Families and Work Institute.
Ganster, D. C. (1989). Measurement of worker control (Final report, Contract No. 88 –79187). Washington, DC: National Institute of Occu- pational Safety and Health.
Gino, F., Schweitzer, M. E., Mead, N. L., & Ariely, D. (2011). Unable to resist temptation: How self-control depletion promotes unethical behav- ior. Organizational Behavior and Human Decision Processes, 115, 191–203.
Glomb, T. M., Bhave, D. P., Miner, A. G., & Wall, M. (2011). Doing good, feeling good: Examining the role of organizational citizenship behaviors in changing mood. Personnel Psychology, 64, 191–223. http://dx.doi .org/10.1111/j.1744-6570.2010.01206.x
Goldberg, D. P., & Hillier, V. F. (1979). A scaled version of the General Health Questionnaire. Psychological Medicine, 9, pp. 139 –145.
Grandey, A. A., Cordeiro, B. L., & Crouter, A. C. (2005). A longitudinal and multi-source test of the work-family conflict and job satisfaction relationship. Journal of Occupational and Organizational Psychology, 78, 305–323. http://dx.doi.org/10.1348/096317905X26769
Grandey, A. A., & Cropanzano, R. (1999). The conservation of resources model applied to work–family conflict and strain. Journal of Vocational Behavior, 54, 350 –370. http://dx.doi.org/10.1006/jvbe.1998.1666
Grant, A. M. (2008). Designing jobs to do good: Dimensions and psycho- logical consequences of prosocial job characteristics. The Journal of Positive Psychology: Dedicated to Furthering Research and Promoting Good Practice, 3, 19 –39.
Grant, A. M., Campbell, E. M., Chen, G., Cottone, K., Lapedis, D., & Lee, K. (2007). Impact and the art of motivation maintenance: The effects of contact with beneficiaries on persistence behavior. Organizational Be- havior and Human Decision Processes, 103, 53– 67. http://dx.doi.org/ 10.1016/j.obhdp.2006.05.004
Grant, A. M. & Gino, F. (2010). A little thanks goes a long way: Explain- ing why gratitude expressions motivate prosocial behavior. Journal of Personality and Social Psychology, 98, 946 –955. http://dx.doi.org/ 10.1037/a0017935
Greenhaus, J., & Beutell, N. J. (1985). Sources of conflict between work and family roles. Academy of Management Review, 10, 76 – 88.
Greenhaus, J. H., & Powell, G. N. (2003). When work and family collide: Deciding between competing role demands. Organizational Behavior and Human Decision Processes, 90, 291–303. http://dx.doi.org/10.1016/ S0749-5978(02)00519-8
Hagger, M. S., Wood, C., Stiff, C., & Chatzisarantis, N. L. D. (2010). Ego depletion and the strength model of self-control: A meta-analysis. Psy- chological Bulletin, 136, 495–525. http://dx.doi.org/10.1037/a0019486
Halbesleben, J. R. B. (2006). Sources of social support and burnout: A meta-analytic test of the conservation of resources model. Journal of Applied Psychology, 91, 1134 –1145. http://dx.doi.org/10.1037/0021- 9010.91.5.1134
Heaphy, E. D., & Dutton, J. E. (2008). Positive social interactions and the human body at work: Linking organizations and physiology. Academy of Management Review, 33, 137–162. http://dx.doi.org/10.5465/AMR .2008.27749365
Heaton, A. W., & Kruglanski, A. W. (1991). Person perception by intro- verts and extraverts under time pressure: Effects of need for closure. Personality and Social Psychology Bulletin, 17, 161–165. http://dx.doi .org/10.1177/014616729101700207
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
789WORK–FAMILY CONFLICT AND TIME ALLOCATION
Heilman, M. E., & Okimoto, T. G. (2007). Why are women penalized for success at male tasks?: The implied communality deficit. Journal of Applied Psychology, 92, 81–92.
Heilman, M. E., Wallen, A. S., Fuchs, D., & Tamkins, M. M. (2004). Penalties for success: Reactions to women who succeed at male gender- typed tasks. Journal of Applied Psychology, 89, 416 – 427. http://dx.doi .org/10.1037/0021-9010.89.3.416
Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological Review, 94, 319 –340. http://dx.doi.org/10.1037/0033- 295X.94.3.319
Higgins, E. T., Bond, R. N., Klein, R., & Strauman, T. (1986). Self- discrepancies and emotional vulnerability: How magnitude, accessibil- ity, and type of discrepancy influence affect. Journal of Personality and Social Psychology, 51, 5–15. http://dx.doi.org/10.1037/0022-3514.51 .1.5
Higgins, E. T., Klein, R., & Strauman, T. (1985). Self-concept discrepancy theory: A psychological model for distinguishing among different as- pects of depression and anxiety. Social Cognition, 3, 51–76. http://dx .doi.org/10.1521/soco.1985.3.1.51
Higgins, E. T., Shah, J., & Friedman, R. (1997). Emotional responses to goal attainment: Strength of regulatory focus as moderator. Journal of Personality and Social Psychology, 72, 515–525. http://dx.doi.org/ 10.1037/0022-3514.72.3.515
Hobfoll, S. E. (1989). Conservation of resources. A new attempt at con- ceptualizing stress. American Psychologist, 44, 513–524. http://dx.doi .org/10.1037/0003-066X.44.3.513
Humphrey, S. E., Nahrgang, J. D., & Morgeson, F. P. (2007). Integrating motivational, social, and contextual work design features: A meta- analytic summary and theoretical extension of the work design literature. Journal of Applied Psychology, 92, 1332–1356. http://dx.doi.org/ 10.1037/0021-9010.92.5.1332
James, W. (1890). The principles of psychology. New York, NY: Dover. http://dx.doi.org/10.1037/11059-000
Judge, T. A., Bono, J. E., Erez, A., & Locke, E. A. (2005). Core self- evaluations and job and life satisfaction: The role of self-concordance and goal attainment. Journal of Applied Psychology, 90, 257–268. http://dx.doi.org/10.1037/0021-9010.90.2.257
Judge, T. A., Cable, D. M., Boudreau, J. W., & Bretz, R. D. (1995). An empirical investigation of the predictors of executive career success. Personnel Psychology, 48, 485–519. http://dx.doi.org/10.1111/j.1744- 6570.1995.tb01767.x
Judge, T. A., Ilies, R., & Scott, B. A. (2006). Work-family conflict and emotions: Effects at work and at home. Personnel Psychology, 59, 779 – 814. http://dx.doi.org/10.1111/j.1744-6570.2006.00054.x
Juhn, C. (1999). Wage inequality and the demand for skill: Evidence from five decades. Industrial & Labor Relations Review, 52, 424 – 443. http:// dx.doi.org/10.2307/2525143
Kahn, R. L., Wolfe, D. M., Quinn, R. P., Snoek, J. D., & Rosenthal, R. A. (1964). Organizational stress: Studies in role conflict and ambiguity. New York: Wiley.
Koestner, R., Lekes, N., Powers, T. A., & Chicoine, E. (2002). Attaining personal goals: Self-concordance plus implementation intentions equals success. Journal of Personality and Social Psychology, 83, 231–244. http://dx.doi.org/10.1037/0022-3514.83.1.231
Kossek, E. E., Lautsch, B. A., & Eaton, S. C. (2006). Telecommuting, control, and boundary management: Correlates of policy use and prac- tice, job control, and work-family effectiveness. Journal of Vocational Behavior, 68, 347–367. http://dx.doi.org/10.1016/j.jvb.2005.07.002
Kristof-Brown, A. L., Zimmerman, R. D., & Johnson, E. C. (2005). Consequences of individuals’ fit at work: A meta-analysis of person-job, person-organization, person-group, and person-supervisor fit. Personnel Psychology, 58, 281–342. http://dx.doi.org/10.1111/j.1744-6570.2005 .00672.x
Kruglanski, A. W. (1989). Lay epistemics and human knowledge: Cogni- tive and motivational bases. New York, NY: Plenum Press. http://dx.doi .org/10.1007/978-1-4899-0924-4
Kruglanski, A. W. (1990). Lay epistemic theory in social cognitive psy- chology. Psychological Inquiry, 1, 181–197. http://dx.doi.org/ 10.1207/s15327965pli0103_1
Leroy, S. (2009). Why is it so hard to do my work? The challenge of attention residue when switching between work tasks. Organizational Behavior and Human Decision Processes, 109, 168 –181. http://dx.doi .org/10.1016/j.obhdp.2009.04.002
Leslie, L. M., Manchester, C. F., Park, T.-Y., & Mehng, S. A. (2012). Flexible work practices: A source of career premiums or penalties? Academy of Management Journal, 55, 1407–1428. http://dx.doi.org/ 10.5465/amj.2010.0651
Liss, M., Schiffrin, H. H., & Rizzo, K. M. (2013). Maternal guilt and shame: The role of self-discrepancy and fear of negative evaluation. Journal of Child and Family Studies, 22, 1112–1119. http://dx.doi.org/ 10.1007/s10826-012-9673-2
Maertz, C. P., & Boyar, S. L. (2011). Work-family conflict, enrichment and balance under “levels” and “episodes” approaches. Journal of Manage- ment, 37, 68 –98. http://dx.doi.org/10.1177/0149206310382455
Maier, G. W., & Brunstein, J. C. (2001). The role of personal work goals in newcomers’ job satisfaction and organizational commitment: A lon- gitudinal analysis. Journal of Applied Psychology, 86, 1034 –1042. http://dx.doi.org/10.1037/0021-9010.86.5.1034
March, J., & Simon, H. (1958). Organizations. New York, NY: Wiley. Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist,
41, 954 –969. http://dx.doi.org/10.1037/0003-066X.41.9.954 Massy, W. F., & Zemsky, R. (1994). Faculty discretionary time: Depart-
ments and the “academic ratchet.” The Journal of Higher Education, 65, 1–22. http://dx.doi.org/10.2307/2943874
Mesmer-Magnus, J. R., & Viswesvaran, C. (2005). Convergence between measures of work-to-family and family-to-work conflict: A meta- analytic examination. Journal of Vocational Behavior, 67, 215–232. http://dx.doi.org/10.1016/j.jvb.2004.05.004
Michel, J. S., Mitchelsen, J. K., Kotrba, L. M., LeBreton, J. M., & Baltes, B. B. (2009). A comparative test of work-family conflict models and critical examination of work-family linkages. Journal of Vocational Behavior, 74, 199 –218. http://dx.doi.org/10.1016/j.jvb.2008.12.005
Milkman, K. L., Minson, J. A., & Volpp, K. G. M. (2013). Holding the hunger games hostage at the gym: An evaluation of temptation bundling. Management Science, 60, 283–299. http://dx.doi.org/10.1287/mnsc .2013.1784
Mischel, W., Cantor, N., & Feldman, S. (1996). Principles of self- regulation: The nature of willpower and self-control. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic prin- ciples (pp. 329 –360). New York, NY: Guilford Press.
Misra, J., Lundquist, J. H., Holmes, E., & Agiomavritis, S. (2011, January/ February). The ivory ceiling of service work. Academe. Available at http://www.aaup.org/article/ivory-ceiling-service-work#.VHOUhct0w90
Moretti, M. M., & Higgins, E. T. (1999). Own versus other standpoints in self-regulation: Developmental antecedents and functional conse- quences. Review of General Psychology, 3, 188 –223.
Morgeson, F. P., & Humphrey, S. E. (2006). The Work Design Question- naire (WDQ): Developing and validating a comprehensive measure for assessing job design and the nature of work. Journal of Applied Psy- chology, 91, 1321–1339. http://dx.doi.org/10.1037/0021-9010.91.6.1321
Muraven, M., & Baumeister, R. F. (2000). Self-regulation and depletion of limited resources: Does self-control resemble a muscle? Psychological Bulletin, 126, 247–259. http://dx.doi.org/10.1037/0033-2909.126.2.247
National Center for Education Statistics. (1999). National Study of Post- secondary Faculty: Faculty questionnaire. Washington, DC: U.S. De- partment of Education Office of Educational Research and Improve- ment.
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
790 DAHM, GLOMB, MANCHESTER, AND LEROY
Netemeyer, R. G., Boles, J. S., & McMurrian, R. (1996). Development and validation of work-family conflict and family-work conflict scales. Jour- nal of Applied Psychology, 81, 400 – 410. http://dx.doi.org/10.1037/ 0021-9010.81.4.400
Obodaru, O. (2012). The self not taken: How alternative selves develop and how they influence our professional lives. Academy of Management Review, 37, 34 –57.
Papies, E. K., & Aarts, H. (2011). Nonconscious self-regulation, or the automatic pilot of human behavior. In K. Vohs & R. F. Baumeister (Eds.), Handbook of self-regulation (2nd ed., pp. 125–142). New York, NY: Guilford Press.
Persell, C. H. (1983). Gender, rewards and research in education. Psychol- ogy of Women Quarterly, 8, 33– 47. http://dx.doi.org/10.1111/j.1471- 6402.1983.tb00615.x
Pocheptsova, A., Amir, O., Dhar, R., & Baumeister, R. F. (2009). Deciding without resources: Resource depletion and choice in context. Journal of Marketing Research, 46, 344 –355. http://dx.doi.org/10.1509/jmkr.46.3 .344
Polasky, L. J., & Holahan, C. K. (1998). Maternal self-discrepancies, interrole conflict, and negative affect among married professional women with children. Journal of Family Psychology, 12, 388 – 401. http://dx.doi.org/10.1037/0893-3200.12.3.388
Powell, G. N., & Greenhaus, J. H. (2006). Managing incidents of work- family conflict: A decision-making perspective. Human Relations, 59, 1179 –1212. http://dx.doi.org/10.1177/0018726706069765
Pratt, M. G. (2000). The good, the bad, and the ambivalent: Managing identification among Amway distributors. Administrative Science Quar- terly, 45, 456 – 493. http://dx.doi.org/10.2307/2667106
Pratt, M. G., & Dutton, J. E. (2000). Owning up or opting out: The role of emotions and identities in issue ownership. In N. Ashkanasy, C. Hartel, & W. Zerbe (Eds.), Emotions in the workplace, research, theory, and practice (pp. 103–129). Westport, CT: Quorum.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple me- diator models. Behavior Research Methods, 40, 879 – 891. http://dx.doi .org/10.3758/BRM.40.3.879
Reichl, C., Leiter, M. P., & Spinath, F. M. (2014). Work-nonwork conflict and burnout: A meta-analysis. Human Relations, 67, 979 –1005. http:// dx.doi.org/10.1177/0018726713509857
Richeson, J. A., & Shelton, J. N. (2003). When prejudice does not pay: Effects of interracial contact on executive function. Psychological Sci- ence, 14, 287–290. http://dx.doi.org/10.1111/1467-9280.03437
Roth, P. L., & BeVier, C. A. (1998). Response rates in HRM/OB survey research: Norms and correlations, 1990 –1994. Journal of Management, 24, 97–117.
Rothbard, N. P. (2001). Enriching or depleting? The dynamics of engage- ment in the work and family roles. Administrative Science Quarterly, 46, 655– 684. http://dx.doi.org/10.2307/3094827
Rothbard, N. P., & Edwards, J. R. (2003). Investment in work and family roles: A test of identity and utilitarian motives. Personnel Psychology, 56, 699 –729. http://dx.doi.org/10.1111/j.1744-6570.2003.tb00755.x
Ryan, R. M., & Deci, E. L. (2006). Self-regulation and the problem of human autonomy: Does psychology need choice, self-determination, and will? Journal of Personality, 74, 1557–1586. http://dx.doi.org/ 10.1111/j.1467-6494.2006.00420.x
Schmeichel, B. J., & Inzlicht, M. (2013). Incidental and integral effects of emotions on self-control. In M. D. Robinson, E. R. Watkins, & E. Harmon-Jones (Eds.), Handbook of cognition and emotion (pp. 272– 290). New York, NY: Guilford Press.
Scott, B. A., & Barnes, C. M. (2011). A multilevel field investigation of emotional labor, affect, work withdrawal, and gender. Academy of Management Journal, 54, 116 –136. http://dx.doi.org/10.5465/AMJ .2011.59215086
Sheldon, K. M., & Elliot, A. J. (1999). Goal striving, need satisfaction, and longitudinal well-being: The self-concordance model. Journal of Per- sonality and Social Psychology, 76, 482– 497.
Shockley, K. M. (2013, April). Fit between desired and actual labor division in dual-earner couples. Symposium presented at the annual Society for Industry and Organizational Psychology conference, Hous- ton, TX.
Shockley, K. M., & Singla, N. (2011). Reconsidering work–family inter- actions and satisfaction: A meta-analysis. Journal of Management, 37, 861– 886. http://dx.doi.org/10.1177/0149206310394864
Smith, P. C., Kendall, L., & Hulin, C. L. (1969). The measurement of satisfaction in work and retirement. Chicago, IL: Rand McNally.
Sonnentag, S., & Fritz, C. (2007). The Recovery Experience Question- naire: Development and validation of a measure for assessing recuper- ation and unwinding from work. Journal of Occupational Health Psy- chology, 12, 204 –221. http://dx.doi.org/10.1037/1076-8998.12.3.204
Stanton, J. M., Sinar, E. F., Balzer, W. K., Julian, A. F., Thoresen, P., Aziz, S., . . . Smith, P. C. (2001). Development of a compact measure of job satisfaction: The abridged Job Descriptive Index. Educational and Psy- chological Measurement, 61, 1104 –1122. Available at http://www .google.com/url?sa�t&rct�j&q�&esrc�s&source�web&cd�1& v e d�0 C C A Q F j A A & u r l�h t t p % 3 A % 2 F % 2 F w w w . r e s e a r c h g a t e .net%2Fprofile%2FJeffrey_Stanton%2Fpublication%2F232533532_ Development_of_a_compact_measure_of_job_satisfaction_The_ abridged_Job_Descriptive_Index%2Flinks%2F00b7d51e029ba3e5a000 0000&ei�5ZdzVJPSIYWmgwSjsoOgDg&usg�AFQjCNGIV8NKa OSO8kiKrw3mCyjsq6OIIw&bvm�bv.80185997,d.eXY
Strauman, T. J. (1996). Stability within the self: A longitudinal study of the structural implications of self-discrepancy theory. Journal of Personality and Social Psychology, 71, 1142–1153. http://dx.doi.org/10.1037/0022- 3514.71.6.1142
Tangney, J. P., Niedenthal, P. M., Covert, M. V., & Barlow, D. H. (1998). Are shame and guilt related to distinct self-discrepancies? A test of Higgins’s (1987) hypotheses. Journal of Personality and Social Psy- chology, 75, 256 –268. http://dx.doi.org/10.1037/0022-3514.75.1.256
Tenbrunsel, A. E., Brett, J. M., Maoz, E., Stroh, L. K., & Reilly, A. H. (1995). Dynamic and static work–family relationships. Organizational- Behavior and Human Decision Processes, 63, 233–246.
Terborg, J., & Miller, H. (1978). Motivation, behavior and performance: A closer examination of goal-setting and monetary incentives. Journal of Applied Psychology, 63, 29 –39. http://dx.doi.org/10.1037/0021-9010.63 .1.29
Thompson, J. A., & Bunderson, J. S. (2001). Work-nonwork conflict and the phenomenology of time: Beyond the balance metaphor. Work and Occupations, 28, 17–39. http://dx.doi.org/10.1177/0730888401028001003
Tice, D. M., Bratslavsky, E., & Baumeister, R. F. (2001). Emotional distress regulation takes precedence over impulse control: If you feel bad, do it! Journal of Personality and Social Psychology, 80, 53– 67. http://dx.doi.org/10.1037/0022-3514.80.1.53
Trougakos, J., Hideg, I., Cheng, B., & Beal, D. (2013). Lunch breaks unpacked: The role of autonomy as moderator of recovery during lunch. Academy of Management Journal. Advance online publication. http:// dx.doi.org/10.5465/amj.2011.1072
Vohs, K. D., Baumeister, R. F., Schmeichel, B. J., Twenge, J. M., Nelson, N. M., & Tice, D. M. (2008). Making choices impairs subsequent self-control: A limited-resource account of decision making, self- regulation, and active initiative. Journal of Personality and Social Psy- chology, 94, 883– 898. http://dx.doi.org/10.1037/0022-3514.94.5.883
Vohs, K. D., & Faber, R. J. (2007). Spent resources: Self-regulatory resource availability affects impulse buying. The Journal of Consumer Research, 33, 537–547. http://dx.doi.org/10.1086/510228
Vohs, K. D., & Heatherton, T. F. (2000). Self-regulatory failure: A resource-depletion approach. Psychological Science, 11, 249 –254. http://dx.doi.org/10.1111/1467-9280.00250
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
791WORK–FAMILY CONFLICT AND TIME ALLOCATION
Wanous, J. P., & Hudy, M. J. (2001). Single-item reliability: A replication and extension. Organizational Research Methods, 4, 361–375. http://dx .doi.org/10.1177/109442810144003
Wanous, J. P., Reichers, A. E., & Hudy, M. J. (1997). Overall job satisfaction: How good are single-item measures? Journal of Applied Psychology, 82, 247–252. http://dx.doi.org/10.1037/0021-9010.82.2.247
Webster, D. M., Richter, L., & Kruglanski, A. W. (1996). On leaping to conclusions when feeling tired: Mental fatigue effects on impressional primacy. Journal of Experimental Social Psychology, 32, 181–195. http://dx.doi.org/10.1006/jesp.1996.0009
Weick, K. E. (1995). Sensemaking in organizations. Thousand Oaks, CA: Sage.
Winslow, S. (2010). Gender inequality and time allocations among aca- demic faculty. Gender & Society, 24, 769 –793. http://dx.doi.org/ 10.1177/0891243210386728
Wood, R. E., Mento, A. J., & Locke, E. A. (1987). Task complexity as a moderator of goal effects: A meta-analysis. Journal of Applied Psychol- ogy, 72, 416 – 425. http://dx.doi.org/10.1037/0021-9010.72.3.416
Wood, W., Christensen, P. N., Hebl, M. R., & Rothgerber, H. (1997). Conformity to sex-typed norms, affect, and the self-concept. Journal of Personality and Social Psychology, 73, 523–535. http://dx.doi.org/ 10.1037/0022-3514.73.3.523
Wooldridge, J. M. (2002). Introductory econometrics: A modern approach. Cincinnati, OH: Cengage Learning.
Received July 18, 2013 Revision received October 21, 2014
Accepted November 3, 2014 �
T hi
s do
cu m
en t
is co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
ti on
or on
e of
it s
al li
ed pu
bl is
he rs
. T
hi s
ar ti
cl e
is in
te nd
ed so
le ly
fo r
th e
pe rs
on al
us e
of th
e in
di vi
du al
us er
an d
is no
t to
be di
ss em
in at
ed br
oa dl
y.
792 DAHM, GLOMB, MANCHESTER, AND LEROY
- Work–Family Conflict and Self-Discrepant Time Allocation at Work
- Derivation of Hypotheses
- Self-Discrepant Time Allocations at Work
- WTF Conflict and Self-Discrepant Time Allocations at Work
- WTF conflict and depletion
- WTF conflict and time allocation
- Self-Discrepant Time Allocation, Work Satisfaction, and Well-Being
- Self-Discrepancy and Salary
- Mediation Hypotheses
- Self-Discrepancy and Gender
- Method
- Participants and Procedure
- Measures
- Time allocations
- Self-discrepancy
- WTF conflict
- Work satisfaction
- Psychological well-being
- Physical well-being
- Annual salary
- Controls
- Analysis
- Results
- Gender results
- Longitudinal Analysis
- Supplemental Analysis: Work Hours and Obligations
- Supplemental Analysis: Academic Rank
- Discussion
- Theoretical and Empirical Contributions
- Gender Effects
- Examination of Alternative Mechanisms
- Implications
- Limitations and Future Directions
- References