RESEARCH REPORTS
Job Characteristics and College Performance and Attitudes:A Model of
Work–School Conflict and Facilitation
Adam B.Butler
University of Northern Iowa
The process linking job characteristics to school performance and satisfaction in a sample of 253full-time college students were examined from 2role theory perspectives,1of which emphasized resource scarcity and the other resource expansion.Model tests using structural equation modeling showed that 2resource-enriching job characteristics,job–school congruence and job control,were positively related to work–school facilitation (WSF).Two resource-depleting job characteristics,job demands and work hours,were positively related to work–school conflict (WSC),and job control was negatively related to WSC.In turn,WSF was positively related to school performance and satisfaction,and WSC was negatively related to school performance.Both WSF and WSC mediated the relationship between the job characteristics and school outcomes.There was no evidence of interactive effects between enriching and depleting job characteristics on interrole process.
office furnitureKeywords:college student employment,work–school conflict,work–school facilitation,college perfor-mance,college students
From 1986to 2004,college tuition and fees ro 240%,whereas textbook prices ro 200%(U.S.Government Accountability Of-fice,2004).During that same period,the percentage of employed full-time college students and students working more than 20hr per week also ro (Fox,Connolly,&Snyder,2005).Among college students who were enrolled in 4-year institutions in 2000,77%were employed,working an average of 27hr per week (U.S.Department of Education,2002).Despite the time commitment required of work,most of the students viewed their work role as less important than their school role.For example,employed students were more likely to classify themlves as “students who work”rather than as “employees who take class”(U.S.Depart-ment of Education,2002),and many reported that they would rather not work during the academic year (Curtis &Williams,2002).Given the primacy of the school role and the economic and social benefits associated with a college education (e.g.,Institute for Higher Education Policy,2005),it is important to identify the characteristics of work that affect school outcomes for college students as well as to understand the process through which such effects occur.
Much student employment rearch is bad on an assumption of resource scarcity in which potenti
al resources such as time,personal energy,and attention are viewed as finite.Using finite resources to fulfill the demands of one role necessitates a reduction
in resource availability for other roles,constraining performance in tho roles.When the demands of one role are incompatible with the demands of another,tension in the form of interrole conflict is experienced (Kahn,Wolfe,Quinn,Snoek,&Ronthal,1964).In their minal work describing interrole conflict between work and family,Greenhaus and Beutell (1985)identified time,strain,and behavior as three forms of interrole conflict.Markel and Frone (1998)similarly suggested that conflict between work and school occurs when work requires time away from school-related activi-ties or when work creates strain that impedes school performance.A number of studies are consistent with the view that working consumes scarce resources that may otherwi be ud to perform the school role.Most of this rearch has been conducted on adolescents and has been limited to an exploration of working per or work hours.In a review of this literature,Frone (1999)concluded that there was mixed evidence of a relationship between employment and academic outcomes,although few studies have reported a positive relationship (e also Mortimer,2003).Other rearchers have concluded,on the basis of the preponderance of evidence,that working negatively affects academic performance for adolescent workers (Steinberg,Fegley,&Dornbusch,1993).Studies of college student
s are largely consistent with this view.Rearch has found that longer hours spent working were associ-ated with poorer study skills (Lammers,Onweugbuzie,&Slate,2001),longer time to graduation (Canabal,1998),and poorer academic performance (Di,1996;Trockel,Barnes,&Egget,2000).Although fewer studies have examined qualitative charac-teristics of student jobs,Barling,Rogers,and Kelloway (1995)found that job role ambiguity was positively related to class cutting by high school students.In sum,a body of rearch suggests that employment may interfere with performance of the school role.
I thank Joe Grzywacz for comments on a draft of this article and Katie Viet,Elissa Narigon,and Emily Taylor for their assistance with study preparation and data collection.
Correspondence concerning this article should be addresd to Adam Butler,Department of Psychology,University of Northern Iowa,Cedar Falls,IA 50614-0505.E-mail:adam.butler@uni.edu
Journal of Applied Psychology Copyright 2007by the American Psychological Association 2007,Vol.92,No.2,500–5100021-9010/07/$12.00DOI:10.1037/0021-9010.92.2.500
500
An alternative to the assumption of resource scarcity,largely overlooked in the student employment literature,is the prospect of resource expansion through multiple role ,Barnett &Hyde,2001;Ruderman,Ohlott,Panzer,&King,2002).Theo-rists have argued that role-bad resources,such as human energy, are not finite and that performance in multiple role domains is beneficial to individuals(Marks,1977;Sieber,1974).Expanding on the ideas,work–family theorists recently suggested that one role domain may offer resources that can be exploited within another domain,leading to interrole facilitation or enrichment (Greenhaus&Powell,2006;Grzywacz,2002).More specifically, Greenhaus and Powell(2006)identified five types of resources that may be accumulated through role occupancy,including skills and perspectives,psychological and physical resources,social capital,flexibility,and material resources.They further suggested that the resources may enrich experiences in another role through instrumental means,enhancing role performance,or af-fective means,enhancing role-related positive emotions.Although the propositions concern interrole facilitation between work and family,similar mechanisms may promote facilitation between work and other roles,such as school.
Although there is not a large body of literature addressing the benefits of work for students,rearch is consistent with the prediction that employment may provide resources that facilitate the school rol
e.For example,many students and their parents report that work is beneficial(Mortimer,2003),and working can promote feelings of mastery and mental health in young workers, attributes that would be expected to benefit school engagement (Finch,Shanahan,Mortimer,&Ryu,1991;Miller,1988;Mor-timer,Harley,&Staff,2002).Similarly,Barling et al.(1995) reported that greater job role clarity,skill variety,and autonomy were all associated with incread time spent studying.In sum, the findings illustrate that working may be beneficial for stu-dents and,more specifically,that certain job characteristics can enhance school outcomes.
hurts是什么意思
A Model of Work–School Conflict and Facilitation Although a number of rearchers have linked characteristics of work to academic outcomes,the role-related mechanisms through which work benefits or harms performance of the school role are poorly understood.In the only study of intervening variables to date,Markel and Frone(1998)reported that interrole conflict between work and school mediated the relationship between job characteristics and negative school outcomes for adolescents.No-tably,there are no studies of the mechanism through which work may have a positive influence on school outcomes.However,a number of rearchers recently have propod that a process of interrole facilitation or enrichment can explain how participation in work benefits the family or vice ,Friedman& Greenhaus,2000;Greenhaus&Powell,2006;Grzywacz,2000; Grz
ywacz&Marks,2000;Kirchmeyer,1992,1993;Voydanoff, 2004a).It ems likely that a similar mechanism of interrole facilitation may explain beneficial relationships between work and school.
At the core of the model prented in Figure1is the proposition that job characteristics are related to school outcomes through their relationship with the quality of the work–school role interface. Following recent work by work–family ,Bakker &Guerts,2004;Voydanoff,2004a,2004b),I propod that some characteristics of student jobs enrich resources,leading to work–school facilitation and better school outcomes,whereas other job characteristics deplete resources,leading to work–school conflict and poorer school outcomes.Work–school conflict(WSC)is de-fined as interference in the school role by work role-related de-mands and responsibilities(Markel&Frone,1998),and work–school facilitation(WSF)is defined as improvement in the quality of the school role resulting from participation in work.(cf.Green-haus&Powell,2006).Although work–family rearchers often asss the bidirectional flow of conflict and facilitation between work and family(Frone,2003),understanding how work affects school is principally important given the primacy of the school role for most students.Becau of a lack of relevant rearch on the interction of work and school roles,some of the model predic-tions prented below,particularly tho involving WSF,were informed by work–family rearch.
Job Characteristics,WSF,and School Outcomes Following theories of resource expansion(Grzywacz,2002; Marks,1977;Sieber,1974),I expected two job characteristics to increa WSF by enriching resources that could be ud to perform the school role.First,I predicted that control over what and how work is done would be related to incread WSF becau it would generate psychological resources(e Greenhaus&Powell,2006) that could be applied in the school domain.The job characteristics model predicts that job control or autonomy generates the psycho-logical resource of responsibility(Hackman&Oldham,1976),and job control has also been linked to the experience of positive emotions(Williams&Alliger,1994).Moreover,veral studies have found a direct relationship between job control and interrole facilitation between work and family(Butler,Grzywacz,Bass,& Linney,2005;Grzywacz&Butler,2005;Grzywacz&Marks, 2000;Voydanoff,2004a,2004b).
Hypothesis1:Greater job control is related to incread WSF.
A cond job characteristic,job–school congruence,was ex-pected to be related to incread levels of WSF.Job–school con-gruence exists when job requirements and collegiate learning are complementary,such that the job requires knowledge or skills acquired in college(cf.Cabel&Edwards,2004;Campbell,1990).
I expected this factor to be related to incread WSF becau it would generate skill and perspective resources(e Greenhaus& Powell,2006)that could be applied in the school domain.For example,job–school congruence may allow students to apply concepts learned in school,increasing conceptual understanding and appreciation for the value of higher education.Consistent with this view,Mortimer(2003)reported that job–school congruence was associated with enhanced learning opportunities,a factor that is related to incread interrole facilitation(Voydanoff,2004a). Moreover,job–school congruence was among the most frequently mentioned positive characteristics of work(Hammes&Heller, 1983),and other studies have found that it was related to incread job satisfaction(Wolniak&Pascarella,2005),enhanced under-standing of vocational interests(Taylor,1988),and decread depressive affect in12th graders(Mortimer et al.,2002).
501
RESEARCH REPORTS
Hypothesis 2:Greater job–school congruence is related to incread WSF.
I hypothesized that job control and job–school congruence would indirectly influence school through both an instrumental path,enhancing performance,and an affective path,enhancing satisfaction (cf.G
盲目的意思reenhaus &Powell,2006).School performance was broadly defined as involvement and demonstrated competence at school,including class attendance,effort,and grades.School satisfaction was defined as attitudes toward the university and educational experiences.Although there are no prior studies of WSF,work–family facilitation has been linked to better physical health,mental health,and well-being (Grzywacz,2000;Grzywacz &Bass,2003).On the basis of the findings,it is reasonable to predict that interrole facilitation would be related to enhanced role performance.Prior work–family studies are also consistent with the view that interrole facilitation is related to more positive role attitudes (Aryee,Srinivas,&Tan,2005);thus,WSF could also be expected to be associated with incread school satisfaction.
Hypothesis 3:More WSF is related to incread school per-formance.
Hypothesis 4:More WSF is related to incread school sat-isfaction.
Hypothesis 5:WSF mediates the relationships of job–school congruence and job control with school performance and satisfaction.
Job Characteristics,WSC,and School Outcomes
On the basis of the theory of resource scarcity,I expected two job characteristics to increa WSC by depleting resources ud to fulfill school role demands.First,I predicted that work hours would be associated with incread WSC becau they deplete a finite resource,time,that might otherwi be devoted to school tasks (cf.Greenhaus &Beutell,1985;Major,Klein,&
Ehrhart,
Figure 1.Conceptual additive model of work–school conflict and facilitation.
502
RESEARCH REPORTS
2002).Consistent with this proposition,veral previous studies of employed adolescents have found that work hours are associated with incread levels of ,Barling et al.,1995;Markel& Frone,1998;Wirtz,Rohrbeck,Charner,&Frar,1988).
Hypothesis6:More hours worked are related to incread WSC.
Second,I predicted that job demands emanating from the pacing and quantity of work would be related to incread WSC becau demands may deplete time for school tasks and increa strain, reducing attention paid to school and personal energy for perfor-mance of school tasks(cf.Greenhaus&Beutell,1985).Consistent with this prediction,Markel and Frone(1998)found that greater workload was associated with higher reported WSC by adoles-cents.Similarly,veral work–family rearchers have reported a positive relationship between job demands and work–family con-
,Aryee et al.,2005;Butler et al.,2005;Frone,Yardley, &Markel,1997;Voydanoff,2004b).
Hypothesis7:Greater job demands are related to incread WSC.
Finally,I expected that one resource-enriching job characteris-tic,control at work,would be negatively associated with WSC. Control over how or what work is completed should increa the permeability of the work role as well as flexibility resources, allowing a college student to fulfill school-related demands at work(cf.Ashforth,Kreiner,&Fugate,2000;Greenhaus&Powell, 2006).Students in jobs allowing for greater control may be able to organize work to discuss school with other employees,lect routine tasks that allow thinking about school,or postpone tasks to accomplish some schoolwork.Barling et al.(1995)reported a marginally significant negative correlation between autonomy and WSC,and a number of work–family rearchers have found that job control was related to lower levels of interrole , Butler et al.,2005;Duxbury,Higgins,&Lee,1994;Grzywacz& Butler,2005;Thomas&Ganster,1995).
Hypothesis8:Greater job control is related to decread WSC.
As with WSF,I also predicted that WSC would indirectly influence school through instrumental means,impairing performance,and af-fective means,decreasing satisfaction(cf.Greenhaus&Powell, 2
006).Prior rearch is consistent with the proposition that WSC will interfere with the performance of effortful behaviors at school,hin-dering school ,Barling et al.,1995;Greenberger, Steinberg,&Vaux,1981;Markel&Frone,1998).Likewi,studies have linked WSC to decread school satisfaction(Hammer,Grigsby, &Woods,1998;Markel&Frone,1998).However,only one study has examined WSC as a mediator between job characteristics and school outcomes(Markel&Frone,1998).
Hypothesis9:More WSC is related to decread school performance.
Hypothesis10:More WSC is related to decread school satisfaction.
Hypothesis11:WSC mediates the relationships of work hours,job demands,and job control with school performance and satisfaction.
Interactive Effects of Job Characteristics on WSF and WSC
Finally,it is possible that enriching and depleting job charac-teristics may interact to affect interrole process.Barling et al. (1995)found that work hours were negatively related to dimen-sions of adolescent school performance only when the quality of employment was low,suggesting that high-q
uality work may buffer the impact of high-intensity work on WSC.There is also a theoretical proposition that job characteristics may interact to promote interrole facilitation.Karak(1979)suggested that jobs high in both demands and control,or active jobs,promote devel-opment and benefit performance outside of work.Although an examination of the impact of active jobs on work–family facilita-tion was not supportive of Karak’s theory(Butler et al.,2005), the proposition has not been subjected to much empirical scrutiny. Thus,job–school congruence and job control may buffer the relationship between demands and work hours and WSC,whereas ,demands and hours)and high-quality , congruence and control)may be associated with enhanced inter-role facilitation.
Hypothesis12:Increasing job–school congruence or job con-trol weakens the relationship between job demands and WSC.
Hypothesis13:Increasing job–school congruence or job con-trol weakens the relationship between work hours and WSC.
Hypothesis14:Increasing job demands or work hours strengthens the relationship between job–school congruence and WSF.
ielts
the time of our lives
Hypothesis15:Increasing job demands or work hours strengthens the relationship between job control and WSF.
Method
Participants
The participants were253employed undergraduate students enrolled full time,averaging14.47mester credit hours(SDϭ1.93),at a4-year residential public university in the United States. More than half of the participants(54.5%)were juniors or niors. The participants reprented38different major areas of study,with education(9%)listed as the most common major.They also worked in a variety of jobs,with most in sales or customer rvice (15%)and food rvice(16%).They worked an average of21.25 hr per week(SDϭ10.40).Sixty-eight percent of the participants were women,the average age was20.75years(SDϭ2.53),and all but20of the participants were White.
Procedure
Participants were solicited through a posting on a departmental rearch recruitment Web site and through in-class announce-ments.Individuals expressing an interest in the study provided
503
RESEARCH REPORTS
their informed connt and were e-mailed a hyperlink to the survey located on a cure Web site.Students responded to questions about their work and school experiences and provided their student identification numbers so that ACT scores(ACT,2006)and grade point averages(GPAs)could be obtained from university records. Measures
Individual scale items are prented in the Appendix.In all cas,items were scored so that higher scores indicated greater levels of the named construct.
Control variables.Age,gender,and ACT composite score were included as model covariates.The demographic variables were lf-reported,and the ACT composite score was obtained from academic records.Seventeen students did not have an ACT score,so to avoid missing data for the participants,I calculated their scores from the linear relationship between ACT and GPA from the remaining sample.
Job–school congruence.Job–school congruence was mea-sured with three items developed for this pantech
study asssing the extent to which learning experiences in college were relevant to work. The items largely reflect the acquisition in college of job perfor-mance determinants,such as declarative and procedural knowl-edge(cf.Campbell,1990).The respon scale for the items was1 (strongly disagree)to5(strongly agree).
fine是什么意思Job control.Job control was measured with three items from the Decision Authority Scale developed by Karak(1979).The items reflected autonomy over decision making and how work is accomplished.The respon scale for the items was1(never/very rarely)to5(always/very often).
Job demands.Job demands were measured with six items from the Job Demands Scale developed by Karak(1979).The items included references to demands such as work quantity, temporal constraints,and conflicts.The respon scale for the items was1(never/very rarely)to5(always/very often).
Hours worked.Participants parately indicated the average number of hours spent in paid labor per week on and off campus. The two figures were summed to give the total number of hours worked per week.
WSF.Five items measuring WSF were developed for this study,modified from a measure of WSF u
d in previous studies (e.g.,Grzywacz&Marks,2000).All of the items reflected facil-itation from work to school.The respon scale for the items was 1(never/very rarely)to5(always/very often).
WSC.WSC was measured with four items developed by Mar-kel and Frone(1998).All items reflected interference from work to school.The respon scale for the items was1(strongly disagree) to5(strongly agree).
School performance.School performance was assd with a range of indicators,including a scale score of school effort,a scale score of school attendance,and GPA in the mester the data were collected,obtained from academic records.The lf-reported school effort measure consisted of nine items,four of which were from the effort scale developed by Markel and Frone(1998)and five of which were developed for this study.The lf-reported attendance measure consisted of four items developed for this study.The respon scale for both the effort and attendance mea-sures was1(never/very rarely)to5(always/very often). School satisfaction.School satisfaction was measured with six items developed for this study.The items reflected satisfaction with being a student,with the educational experience,and with the university in general.The respon scale for the items was1 (strongly disagree)to5(strongly agree).
Resultsupdate rvice
The propod structural equations model,prented in Figure1, was tested using EQS6.1.The substantive model included10 latent variables and was tested both including and excluding three exogenous latent covariates.Several of the latent variables had single-item ,age,gender,academic achievement, and work hours).Given that the constructs were relatively con-crete and either easily known by the participants or obtained from objective indicators,I fixed factor loadings at1.0and measure-ment errors to0.0(Markel&Frone,1998).Descriptive statistics, correlations,and reliability coefficients are prented in Table1. As shown in Table2,all of the measured variables were signifi-cantly related to their respective latent constructs.
Model fit was first assd by comparing the conceptual model excluding covariates to the conceptual model including covariates. The covariates were treated as exogenous latent variables with paths to each of the substantive latent variables in the model. Women had better school performance(␥ϭ–.22)and greater school satisfaction(␥ϭ–.16)than men.Being older was associ-
Table1
Descriptive Statistics,Correlations,and Reliability Coefficients
Variable M SD12345678910
1.Job–school congruence
unenthusiastic2.99 1.06.87
2.Job control
3.150.86.32.85
3.Job demands 2.880.67.13.07.81
4.Work hours21.2010.43.06.03.30—
5.Work–school facilitation 2.860.84.65.39.17.11.85
6.Work–school conflict 2.780.94Ϫ.06Ϫ.2
7.2
8.36Ϫ.15.88
7.School effort 2.530.64.22.12.05Ϫ.03.26Ϫ.17.88
good luck
8.School attendance 4.120.86.13.10Ϫ.01Ϫ.05.23Ϫ.11.61.86
9.Grade point average 2.970.69.08.07Ϫ.09Ϫ.17.11Ϫ.03.41.41—
10.School satisfaction 3.840.82.25.18.04.02.37Ϫ.09.19.19.12.95 Note.Cronbach’s alpha reliability coefficients are on the diagonal.
rϾ.13,pϽ.05.rϾ.17,pϽ.01.
504RESEARCH REPORTS
