The Social Science Journal 50 (2013) 408–417

Contents lists available at ScienceDirect

The Social Science Journal

j ourna l ho me pa g e: www.elsev ier .com/ locate /sosc i j

Workplace characteristics, work-to-life conflict, andpsychological distress among medical workers

Krista Lynn Minnottea,∗, Marissa Gravelleb, Michael C. Minnottec,1

a University of North Dakota, Department of Sociology, 225 Centennial Drive, Stop 7136, Grand Forks, ND 58202 7136, USAb University of North Dakota, Department of Sociology, 225 Centennial, Stop 7136, Grand Forks, ND 58202 7136, USAc University of North Dakota, Department of Mathematics, 101 Cornell, Stop 8376, ND 58202 8376, USA

a r t i c l e i n f o

Article history:Received 28 September 2012Received in revised form 5 August 2013Accepted 5 August 2013Available online 24 August 2013

Keywords:

a b s t r a c t

This study addresses whether work-to-life conflict mediates the relationships betweenworkplace characteristics and psychological distress for workers in the medical indus-try who experience heightened stressors in the workplace. Drawing on data from the2002 National Study of the Changing Workforce, hypotheses are tested using stepwise OLSregression. Overall, the findings indicate that work-to-life conflict mediates the relation-ship between job pressure and supervisor support and psychological distress. The finding’simplications and suggestions for future research are discussed.

Work and familyWork-to-life conflictPsychological distressMedical workerWork-to-family conflict

© 2013 Western Social Science Association. Published by Elsevier Inc. All rights reserved.

1. Introduction

The medical industry is the fastest growing job sector,and it is projected to be the largest job sector in the U.S.by 2018 (Woods, 2009). In the near future, there will likelybe more people with medical needs than there are medi-cal workers to attend to those needs (Dawson, 2012). As aresult, there is a need for more doctors, nurses, and othermedical workers to take care of the healthcare necessitiesof the U.S. population (Johnson, 2010). However, ensuringthat the number of medical workers matches such growth

is hampered by issues of short job tenure and job burnoutamong medical workers that arise from the heightenedworkplace stressors endemic to this industry (Ford, 2013;

∗ Corresponding author at: University of North Dakota, SociologyDepartment, 202 Gillette Hall, 225 Centennial Drive, Stop 7136, GrandForks, ND 58202-7136, USA. Tel.: +1 701 777 4419; fax: +1 701 777 4767.

E-mail addresses: krista.minnotte@und.edu, klminnotte@yahoo.com(K.L. Minnotte).

1 Tel.: +1 701 777 4600.

0362-3319/$ – see front matter © 2013 Western Social Science Association. Publihttp://dx.doi.org/10.1016/j.soscij.2013.08.001

Linzer et al., 2002; Qidwai, Beasley, & Gomez-Clavelina,2008). Given these issues, previous research links work-place stressors to psychological distress, and other relatedmental-health outcomes, among medical workers (Burke &Greenglass, 1999; Parikh, Taukari, & Bhattacharya, 2004).We contribute to this literature by examining whetherworkplace characteristics primarily impact psychologicaldistress among medical workers by increasing work-to-life conflict. In doing so, we highlight how characteristicspresent in the industry at large, rather than just among spe-cific medical occupations, shape the psychological distressof medical workers.

Previous scholarship indicates that employees, ingeneral, experience psychological distress as a resultof work-to-life conflict in which workplace demandsinterfere with personal/family responsibilities (Burke &Greenglass, 1999; Craig de Silva et al., 2008). Studies

also point to work-to-life conflict playing a mediatingrole between work stressors and psychological distressor other mental health outcomes among workers (Grant-Vallone & Ensher, 2001; Hämmig, Gutzwiller, & Bauer,

shed by Elsevier Inc. All rights reserved.

l Scienc

2spefliJManasaiaositl

tcmsthtff2llttma(1lct

2

iaitomaeppts(S

K.L. Minnotte et al. / The Socia

009). Very little research, however, considers relation-hips between work stressors, work-to-life conflict, andsychological distress among medical workers. The fewxisting studies provide evidence that work-to-family con-ict may act as a mediating variable for specific occupations

n the medical industry (Geurts, Rutte, & Peeters, 1999;anssen, Peeters, de Jonge, Houkes, & Tummers, 2004;

ontgomery, Panagopolou, & Benos, 2006). This studydds to these existing studies by incorporating a moreuanced treatment of workplace social support that sep-rately looks at the role of both coworker and supervisorupport and by including important control variables. Welso take a broader contextual approach by highlight-ng the presence of such issues in the medical industryt large rather than limiting our attention to specificccupations in the medical industry. It is important totudy specific occupational groups; however, broadly look-ng at workers across the medical industry allows uso see how these issues operate at a larger contextualevel.

The overall goal of this study is to examine the rela-ionships between workplace characteristics, work-to-lifeonflict, and psychological distress among workers in theedical industry. It is important to explore these relation-

hips among medical workers due to the work stressorshat they face, including long and often non-traditionalours, high job demands, low workplace social support, lit-le job autonomy, and high job pressure levels stemmingrom working directly with people who often have pro-ound problems (Rees, 1995; Roopalekha, Latha, & Swetha,012). These workplace stressors contribute to work-to-

ife conflict among medical workers, which may thenead to mental health issues, such as psychological dis-ress (Geurts et al., 1999). Identifying mechanisms leadingo work-to-life conflict and psychological distress among

edical workers may help address issues of job burnoutnd short job tenure that are evident in this industryAiken & Sloane, 1997; Bourbonnais, Comeau, & Vézina,999; Shields & Ward, 2001). This study contributes to the

iterature by identifying mechanisms contributing to psy-hological distress among a national sample of workers inhe medical industry in the U.S. (N = 246).

. Theoretical framework

Tiedje et al. (1990, p. 64) describe role theory as “view-ng [the] energies of individuals as finite and role demandss infinite . . . [such that] role conflict, then, becomes annevitable, normal, and expected consequence of mul-iple roles”. Role theory’s basic premise is that peopleccupy many different roles at any given time, whichakes role conflict likely. Such role conflict is described

s “a stressful situation that results from discrepant rolexpectations and from the inability to resolve those incom-atible expectations” (Pomaki, Supeli, & Verhoeven, 2007,. 317). Experiencing role conflict is connected to nega-

ive outcomes, such as psychological distress, individualtress, reduced life satisfaction, depression, and anxietyAmstad, Meier, Fasel, Elfering, & Semmer, 2011; Hill, 2005;chieman & Glavin, 2011).

e Journal 50 (2013) 408–417 409

Most scholarship focuses on the role conflict that occursbetween work and family roles, as these two domainsare where many adults spend their time (Hill, 2005;Voydanoff, 2002). Such conflict is viewed as taking threeprimary forms: time-based, strain-based, and behavior-based conflict (Greenhaus & Beutell, 1985). Time-basedconflict occurs when the time obligations of one role inter-fere with the ability to perform the other role, whereasstrain-based conflict stems from stressors in one domainmaking it difficult to fulfill obligations in the other domain.Behavior-based conflict takes place when the patterns ofbehavior expected of one role are incompatible with thepatterns of behavior expected of the other role. We take abroad perspective by studying work-to-life conflict, whichis defined as occurring when demands at work conflictwith demands and responsibilities in one’s personal life(Hämmig et al., 2009). Work-to-life conflict is viewed asstemming primarily from work stressors and can lead todetrimental outcomes within the personal domain. Med-ical workers may be more at risk than workers in otherindustries of experiencing work-to-life conflict due to theirworkplace stressors, including long hours and high jobpressure levels (Burke & Greenglass, 1999; Pisarski et al.,2006; Rees, 1995).

3. Previous literature and hypotheses

3.1. Psychological distress

In this study, psychological distress refers to feelingsof depression, lack of interest in normal activities, feel-ings of nervousness, minor health problems, and problemssleeping (Schieman & Glavin, 2011). Previous scholarshipsuggests that the most commonly reported workplacecharacteristics contributing to psychological distress—andother negative mental health outcomes—among the gen-eral working population include work-to-life conflict, jobpressure, lack of job autonomy, working long hours, non-standard work hours, and low coworker and supervisorsupport (Haines, Marchand, Rousseau, & Demers, 2008;Hämmig et al., 2009; Hughes & Parkes, 2007; Parasuraman& Simmers, 2001; Thompson & Prottas, 2005). Scholarshipalso shows that work-to-life conflict may mediate the rela-tionships between work characteristics and mental healthoutcomes, such as psychological distress, among the gen-eral population (Thompson & Prottas, 2005; Parasuraman,Purohit, Godshalk, & Beutell, 1996). Few studies explorewhether work-to-life conflict mediates the relationshipbetween work characteristics and psychological distressamong medical workers, and those that do exist pointto work-to-life conflict operating as a mediating variable(Geurts et al., 1999; Janssen et al., 2004; Montgomery et al.,2006). Therefore, a mediating conceptual model is pro-posed for medical workers in this study, which is depictedin Fig. 1. The model shows the proposed relationshipsbetween workplace characteristics, work-to-life conflict,

and psychological distress among medical workers, sug-gesting both direct and indirect relationships, and positingwork-to-life conflict as a mediating variable between workcharacteristics and psychological distress

410 K.L. Minnotte et al. / The Social Science Journal 50 (2013) 408–417

WORK PLACE CHARACTERISTICS

Job pr essure

Job autonomy

Work ho urs

Nonstandard work ho urs

Coworker s upport

Supe rvisor support

WORK-TO-LIFE CONFLICT

PSYCHOLOGICAL DISTRES S

orkplac

Fig. 1. Conceptual model of direct and indirect relationships between wmedical workers.

3.2. Workplace characteristics and psychological distress

As workers navigate their work and personal lives,they encounter both demands and resources that shapetheir experiences. Demands create stressors for work-ers, whereas resources may be drawn upon by workersto enhance their well-being (Voydanoff, 2007). Here weconceptualize several workplace characteristics that aresalient in the medical industry as either demands orresources. The demands considered include job pressure,work hours, and nonstandard work hours; resources arejob autonomy, coworker support, and supervisor support.Turning first to demands, Rees (1995) suggests that medicalworkers, as a result of working directly with patients, oftenexperience high job pressure levels. In this study, job pres-sure refers to “the degree to which work and time urgencydominate the work milieu” (Fielding & Weaver, 1994, p.1199). In general, previous research finds that job pres-sure, which is sometimes referred to as job demands orworkload, is associated with poor mental health, includingpsychological distress, stress, and emotional exhaustion(Diestel & Schmidt, 2009; Hill, 2005; Hughes & Parkes,2007; Ilies & Dimotakis, 2010; Linzer et al., 2002; Rout,2000).

Another demand encountered in the workplace is longwork hours, and previous work suggests that long workhours are a risk factor associated with psychological dis-tress and stress (Hilton et al., 2008; Linzer et al., 2002).Among medical workers, we view work hours as a demandbecause they represent greater exposure to the stressfulcharacteristics endemic to this industry. Findings regardingwork hours and psychological distress point to both directand indirect relationships, with one meta-analysis show-ing a direct positive relationship between work hours andthe presence of poor mental health symptoms (Sparks,Cooper, Fried, & Shirom, 1997), and other studies suggest-ing work hours impact psychological distress by shapingwork-to-life conflict (Hämmig et al., 2009; Major, Klein, &Ehrhart, 2002). Altogether, these studies show that overalltime spent at work may be important in predicting psy-chological distress. In addition to number of hours worked,the timing of such hours may also impact psychologi-cal distress. In this study, nonstandard work hours referto work timing that deviates from the standard Monday

through Friday beginning work in the morning and endingwork in the early evening time frame, including workingrotating shifts, working weekends, and working at night(Presser, 2003). Some studies find that nonstandard hours

e characteristics, work-to-life conflict, and psychological distress among

are related to high work-to-life conflict, which then con-tributes to poor mental health outcomes, such as anxietyand depression (Haines et al., 2008; Hämmig et al., 2009).In terms of how nonstandard hours might directly con-nect to psychological distress, studies show such hoursare associated with increased psychological distress, higherworker burnout, disturbed sleep, fatigue, and poor mentalhealth (Akerstedt, Fredlund, Gillberg, & Jansson, 2002; Bildt& Michelsen, 2002; Fenwick & Tausig, 2001; Shields, 2002).Medical workers’ work schedules are often variable andinflexible (Pisarski et al., 2006); hence, this group may belikely to experience poor mental health due to nonstandardwork hours. In line with past research regarding job pres-sure, work hours, and nonstandard work hours, we proposeHypothesis One:

H1. Job pressure, work hours, and nonstandard workhours are positively related to psychological distressamong medical workers.

We also consider how the workplace resources jobautonomy, coworker support, and supervisor supportrelate to the psychological distress of medical workers.Job autonomy is defined here as “the degree to whichthe job provides substantial freedom, independence, anddiscretion to the individual in scheduling work and indetermining the procedures to be used in carrying it out”(Wilson, DeJoy, Vandenberg, Richardson, & McGrath, 2004,p. 572). Having such latitude to structure one’s work maylead workers to acquire a sense of freedom over theirwork environment, which may limit psychological distress.As such, numerous studies link job autonomy or havingperceived control over one’s work environment to positivemental health outcomes, including enhanced psycholog-ical well-being and lower levels of stress (Escribà-Agüir& Pérez-Hoyos, 2007; Pisarski et al., 2006; Thompson &Prottas, 2005).

Two important types of workplace socialsupport—coworker support and supervisor support—mayalso come into play. Coworker support is defined as “infor-mal social/interpersonal relationships that develop amongpeers [in the workplace]” (Wilson et al., 2004, p. 571), andsupervisor support encompasses both support in workers’job-related functions and their desires to “. . .seek balancebetween work and family responsibilities” (Thomas &

Ganster, 1995, p. 7). Research shows coworker supportand supervisor support are related to better mental healthoutcomes, such as increased psychological well-beingand reduced emotional exhaustion (Escribà-Agüir &

l Scienc

PJap

Hsd

isnia&McbsddHPmsfiihp

Haw

3

miatwebu(ietbripisaewoGo

K.L. Minnotte et al. / The Socia

érez-Hoyos, 2007; Pisarski et al., 2006; ter Doest & deonge, 2006). In line with the literature reviewed on jobutonomy, coworker support, and supervisor support, weropose Hypothesis Two:

2. Job autonomy, coworker support, and supervisorupport will be negatively associated with psychologicalistress among medical workers.

Previous research indicates that work-to-life conflicts an important predictor of mental health outcomes. Inome cases, work-to-life conflict is directly associated withegative mental health outcomes, such as depression, anx-

ety, and stress (Grant-Vallone & Ensher, 2001; Hill, 2005),nd in others, it plays a mediating role (Ford, Heinen,

Langkamer, 2007; Kinnunen, Geurts, & Mauno, 2004;ajor et al., 2002; Pisarski et al., 2006), in which work

haracteristics indirectly influence psychological distressy increasing work-to-life conflict. In either case, researchhows that work-to-life conflict increases psychologicalistress, or other negative mental health outcomes, amongifferent types of workers (Grant-Vallone & Ensher, 2001;ämmig et al., 2009; Hill, 2005; Kinnunen et al., 2004;isarski et al., 2006), but research in this area regardingedical workers is limited. We formulate two hypothe-

es about work-to-life conflict because previous researchnds evidence of both a direct relationship and a mediat-

ng relationship for other types of workers. We propose aypothesis predicting an indirect association, but first, weropose Hypothesis three predicting a direct association:

3. Work-to-life conflict will be directly and positivelyssociated with psychological distress among medicalorkers.

.3. Work-to-life conflict and psychological distress

Some studies point to work-to-life conflict acting as aediator of relationships between job characteristics and

ndividual outcomes (Voydanoff, 2002, 2007). Montgomerynd colleagues (2006) posit that the definition of work-o-life conflict theoretically implies mediation, in thatork-to-life conflict “can mediate the way we experi-

nce demands over a prolonged period” (p. 204), andecause work-to-life conflict cannot occur if the work sit-ation does not contain job demands in the first placeJanssen et al., 2004). Hence, we view work character-stics as indirectly impacting psychological distress byither increasing or decreasing work-to-life conflict. Fur-hermore, the mediating model is theoretically supportedy role theory, as its primary assumption centers onole conflict stemming from work stressors contribut-ng to negative outcomes among workers. Indeed, a fewast studies regarding specific occupations in the med-

cal industry provide support for work-to-life conflicterving as a mediating variable between job demandsnd psychological outcomes (Geurts et al., 1999; Janssent al., 2004; Montgomery et al., 2006). These past studies

arrant in-depth consideration of their findings because

f their centrality to this study. The earliest study byeurts and colleagues (1999) examines medical residentsf an academic hospital in the Netherlands finding that

e Journal 50 (2013) 408–417 411

unfavorable work schedules (having low satisfaction withone’s work schedule), high quantitative workloads (work-ing in a high pressure environment), and dependencyon one’s supervisor are associated with higher levels ofwork–home interference, which in turn predict psycho-logical health indicators. Janssen and colleagues (2004)explore similar relationships among U.S. and Dutch nursesworking in nursing homes, with their findings show-ing that negative work–home interference mediates therelationships between psychological job demands (jobpressure), support (among Dutch sample only), job con-trol (among Dutch sample only), and emotional exhaustion.Lastly, Montgomery and colleagues (2006) explore howjob demands and work–family interference relate to emo-tional exhaustion and depersonalization. Their findingssuggest that work–family interference partially mediateshow emotional job demands are associated with deper-sonalization. Altogether, these previous studies underscorework-to-life conflict’s importance as a mediating variableamong various occupational groups in the medical indus-try.

These studies are important in providing an empiricalfoundation for the mediating hypothesis, but we build uponthis previous work in three key ways. First, these existingstudies focus on particular occupations (doctors, medicalresidents of an academic hospital, nurses, and nurse assis-tants), whereas we take a broader approach by exploringvarious occupations across the medical industry, therebyhighlighting how work–life variables operate at this largercontextual level. Second, we provide a more nuanced treat-ment of social support in the workplace by considering theroles of coworker support and supervisor support sepa-rately, as they may uniquely shape psychological distress.One study (Janssen et al., 2004) contains an overall measureof social support in the workplace without looking specif-ically at the different types of support that exist, whereasthe other two studies do not consider workplace social sup-port (Geurts et al., 1999; Montgomery et al., 2006). Third,we incorporate important control variables, such as race,gender, education, and income that are not included in theprevious studies, which lend greater credence to our find-ings. Given both the theoretical and empirical foundationsfor work-to-life conflict serving as a mediating variable, wepropose the following hypothesis:

H4. Work-to-life conflict will mediate the relation-ship between workplace characteristics (job pressure, jobautonomy, work hours, nonstandard work hours, coworkersupport, and supervisor support) and psychological dis-tress among medical workers.

3.4. Control variables and psychological distress

In addition to the primary independent variables,we integrate several control variables into our analyses,including age, race, gender, education, household income,and presence of children under the age of 18 in the

home. We include age because people experience differ-ent work–life challenges across the life course (Erickson,Martinengo, & Hill, 2010), and we include race because itmay shape the stressors faced in the workplace (Minnotte,

l Scienc

412 K.L. Minnotte et al. / The Socia

2012a). We control for gender because men and womenmay experience work and personal life domains differently,and women typically have higher psychological distresslevels than men (Mirowsky & Ross, 1995). Education isimportant to include because previous research suggestseducation shapes how work–family conflict relates to out-comes, such as psychological distress (Schieman & Glavin,2011). On a similar note, we include income as those withhigher incomes may have higher levels of work-to-familyconflict levels and greater resources to address the work-to-life conflict they do experience (Schieman & Glavin,2011). Presence of at least one child under age 18 in thehome is taken into account because such children may con-tribute to work-to-life conflict.

4. Methods

4.1. Data and sample

We use data from the 2002 National Study of the Chang-ing Workforce (NSCW) to test the hypotheses. The NSCW,a telephone survey for working adults aged 18 or over inthe U.S., was initiated by the Families and Work Institute(Thompson & Prottas, 2005). A random digit dialing methodwas used to obtain the sample with a response rate of 52%(Thompson & Prottas, 2005). All participants who did notidentify as medical workers were eliminated from the anal-ysis, leaving a total of 246 respondents after cases withmissing values were deleted. To clarify, this study examinesthe hypotheses among workers in the larger medical indus-try, rather than looking at specific occupational groups,such as doctors or nurses. An industry refers to the differentlarge sectors of the economy ranging from manufactur-ing to financial activities. People who work in the sameindustry may perform different occupations, but they allcontribute to the same area of the economy. By focusingon the medical industry this study examines a wide rangeof medical occupations ranging from those having directcontact with patients to those with little to no contact withpatients.

4.2. Dependent variable

4.2.1. Psychological distressPsychological distress was measured with five items

(Schieman & Glavin, 2011; Voydanoff, 2005). The first threeitems asked respondents “In the last month, how oftenhave you: (1) been bothered by minor health problemssuch as headaches, insomnia, or stomach upsets?; (2) Hadtrouble sleeping to the point that it affected your per-formance on and off the job?; and (3) Felt nervous andstressed?” Response choices are “never” (1), “almost never”(2), “sometimes” (3), “fairly often” (4), and “often” (5). Thelast two items asked respondents, “During the last monthhave you (4) been bothered by feeling down, depressed, orhopeless? and (5) been bothered by little interest or plea-

sure in doing things?” For these last two items, responsechoices are “no” (1) or “yes” (5). The coding of these lasttwo items follows the pattern used by Schieman and Glavin(2011) and Voydanoff (2005). All of the items are averaged,

e Journal 50 (2013) 408–417

with higher scores indicating greater levels of psychologi-cal distress ( = .74).

4.3. Independent variables

4.3.1. Work-to-life conflictWork-to-life conflict was measured with five items

(Minnotte, 2012a, 2012b; Hill, 2005). The response formatranged from “never” (1) to “frequently” (4). The items are:“How often have you not had enough time for your fam-ily or other important people in your life because of yourjob?”; “How often have you not had enough energy to dothings with your family or other important people in yourlife because of your job?”; How often have you not been ina good mood at home because of your job?”; “How oftenhas work kept you from doing as good a job at home as youcould?”; and “How often has your job kept you from con-centrating on important things in your family and personallife?” Responses are summed and divided by five for easeof interpretation. Higher scores indicate a higher level ofwork-to-life conflict, and the scale has an alpha reliabilitycoefficient of .87.

4.3.2. Coworker supportCoworker support is measured with a four-item scale

(Minnotte, 2012a). The items are: “I feel part of the groupof the people I work with”; “I have the coworker supportI need to do a good job”; “I am treated with respect atwork”; and “I have the coworker support I need to managework/family life”. Responses range from “strongly agree”(1) to “strongly disagree” (4). Items are reverse-coded,summed, and averaged, such that higher scores indicatehigher levels of coworker support ( = .75).

4.3.3. Supervisor supportSupervisor support is measured with a nine-item scale

that has been used by previous researchers (Minnotte,2012a; Beutell, 2010). Items are: “My supervisor supportsme when I have a work problem”; “I feel comfortable bring-ing up family and personal matters with my supervisor”;“My supervisor keeps me informed of things I need to domy job well”; “My supervisor has realistic expectations ofmy job performance”; “My supervisor recognizes when Ido a good job”; “My supervisor is fair when responding toemployee personal/family needs”; “My supervisor accom-modates me when I have family/personal business”; “Mysupervisor is understanding when I have family/personalbusiness”; and “My supervisor cares about effects of workon personal/family life”. Response categories range from“strongly agree” (1) to “strongly disagree” (4). The itemsare reverse-coded, summed, and averaged such that a highscore indicates a high level of supervisor support ( = .90).

4.3.4. Job autonomyJob autonomy is measured using three items (Minnotte,

2012a; Schieman & Glavin, 2011; Voydanoff, 2005). Partic-ipants were asked the extent to which they agreed with

the following statements: “I have the freedom to decidewhat I do on my job”; “It is basically my own responsibil-ity to decide how my job gets done”; and “I have a lot ofsay about what happens on my job”. The response format

l Science Journal 50 (2013) 408–417 413

rTi

4

hTnojjrThai

4

attsp

4

atgcmutyGmrcan$Fecdb(

4

srutdnsc

Table 1Descriptive statistics (N = 246).

Variables M S.D. Scale range

Psychological distress 2.18 .51 1–5Work-to-life conflict 2.52 .89 1–5Job pressure 2.86 .70 1–4Job autonomy 3.05 .75 1–4Work hours 43.61 15.58Nonstandard hours .37Coworker support 3.48 .62 1–4Supervisor support 3.34 .67 1–4Age 43.01 13.39Race .79Gender .25Child under 18 .48Education 3.16 1.05Household income 2.05 .68

Source: The National Study of the Changing Workforce 2002.Note. Nonstandard hours is a dummy variable coded 1 if the respondentworks nonstandard hours at their job. Race is a dummy variable that wascoded 1 if the respondent is white and 0 if the respondent is non-white.

K.L. Minnotte et al. / The Socia

anges from “strongly disagree” (1) to “strongly agree” (4).he scores are summed and averaged with higher scoresndicating higher levels of job autonomy ( = .71).

.3.5. Job pressureJob pressure is measured using a five-item scale that

as been used in past research (Schieman & Glavin, 2011).he items are: “My job requires that I work very fast”; “Iever seem to have enough time to get everything donen my job”; “My job requires that I work very hard”; “Myob is very emotionally demanding and tiring”; and “Myob is very physically demanding and tiring.” Responsesanged from “strongly agree” (1) to “strongly disagree” (4).he items are reverse-coded, summed, and averaged withigher scores indicating more pressure. The scale has anlpha reliability coefficient of .69 for those in the medicalndustry.

.3.6. Nonstandard work hoursNonstandard work hours is measured with one item that

ssessed whether participants worked a standard Mondayhrough Friday day shift (coded 0) or another type of shifthat is not a regular day shift (coded 1). Work hours is mea-ured in terms of the number of hours respondents workeder week at all jobs.

.3.7. Control variablesThis study takes into account several demographic vari-

bles that may impact psychological distress. These includehe presence of children under age 18 in the home, age,ender, race, household income, and education. Presence ofhildren was also coded as a dummy variable for which 0eans that the participant does not have at least one child

nder 18 years of age living in the home, and 1 means thathe participant had at least one child under the age of 18ears of age living in the home. Age is measured in years.ender and race are coded as dummy variables. For gender,en are coded as 1 and women are coded as 0. To code

ace, whites are coded 1 and all other racial groups wereoded as 0. Household income and education are measuredt the ordinal level. Respondents’ household income is orga-ized into 3 groups: (1) less than or equal to $27,999, (2)28,000–$79,999, and (3) greater than or equal to $80,000.or the education variable, respondents chose the high-st level of education they completed with the followinghoices: less than a high school degree (1), a high schoolegree or GED (2), completed some college (3), obtained aachelor’s degree (4), or obtained a post-graduate degree5).

.4. Analytic strategy

First, we present descriptive statistics to provide aummary of the sample. Second, we provide bivariate cor-elations for the variables used in the study. Then, wese stepwise ordinary least squares (OLS) regression toest the hypotheses. The regression analysis first tests the

irect relationships between job pressure, work hours,onstandard work hours, job autonomy, coworker support,upervisor support, the control variables, and psychologi-al distress (Model 1). Next, work-to-life conflict is added

Gender is a dummy variable coded 0 if the respondent is female and 1 ifmale. Child under 18 is a dummy variable coded 1 if the respondent has achild under 18 living in the home and 0 if not.

to Model 2 to test the fourth hypothesis, which posits work-to-life conflict as a mediating variable.

In order to establish mediation, four requirements mustbe met (Baron & Kenny, 1986). The first requirement ismet as long as the independent variables (job pressure,job autonomy, work hours, nonstandard work hours, avail-ability of flexible scheduling, coworker support, supervisorsupport,) are significantly related to the mediating vari-able (work-to-life conflict). We test this by performinga separate analysis in which the workplace characteris-tics are used to predict work-to-life conflict. The secondrequirement is met if the significant relationships betweenworkplace characteristics and psychological distress in thefirst model become less significant or non-significant inthe second model once work-to-life conflict is taken intoaccount. The third requirement is met if the mediatingvariable and the dependent variable yield a significantrelationship. The final requirement is met if there is a sig-nificant increase in the percentage of variance explained bythe independent variables with the addition of the medi-ating variable. In addition to the requirements presentedby Baron and Kenny (1986), we also use the Sobel test tofurther test the significance of any mediating relationships(Preacher & Hayes, 2004).

5. Results

5.1. Descriptive statistics and correlations

Descriptive statistics for the variables are shown inTable 1. On average, medical workers report working43.61 h per week (S.D. = 14.58), and 37% work nonstandardhours. The average age of medical workers in the sampleis 43 years (S.D. = 13.39), with 75% of the sample consist-

ing of women and 25% of the sample consisting of men.Forty-eight percent of respondents have at least one childunder the age of 18 living in the home, and the majorityof respondents (79%) are white. The mean for household

414 K.L. Minnotte et al. / The Social Science Journal 50 (2013) 408–417

Table 2Zero-order correlation matrix of variables used in the analysis (N = 246).

Variables X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12 X13 X14

X1: Psychological distress –X2: Work-to-life conflict .46*** –X3: Job pressure .26*** .47*** –X4: Job autonomy −.13* −.14* −.18*** –X5: Work hours .05 .30*** .26*** .06 –X6: Nonstandard hours −.001 −.001 .10 −.17** .01 –X7: Coworker support −.17** −.32*** −.18** .25*** −.08 −.14* –X8: Supervisor support −.27** −.40*** −.26*** .24** −.14* −.19** .52*** –X9: Age −.23*** −.14* −.01 .16** −.02 −.03 .004 .09 –X10: Race .07 .02 .05 .15** −.08 −.04 .17** .10 .21*** –X11: Gender −.13* .04 −.05 .16** .21*** −.04 −.06 −.02 .16** .08 –X12: Child under 18 .02 .10 .08 −.11 .02 .06 .01 −.02 −.33*** −.10 −.09 –X13: Education −.08 .09 .06 .18** .12* −.08 .01 −.05 .12* .16** .29*** −.14* –X14: Household income −.07 .13* .16** .15** .21*** −.08 −.05 −.06 .25*** .17** .16** .01 .38*** –

Source: The National Study of the Changing Workforce 2002.Note. Nonstandard hours is a dummy variable coded 1 if the respondent works nonstandard hours. Race is a dummy variable that was coded 1 if therespondent is white and 0 if the respondent is non-white. Gender is a dummy variable coded 0 if the respondent is female and 1 if male. Child under 18 isa dummy variable coded 1 if the respondent has a child under 18 living in the home and 0 if not.

* p < 0.05.** p < 0.01.

*** p < 0.001.

income is 2.05 (S.D. = .68), which means that on averagerespondents report household incomes between $28,000and $79,999. The average educational level is some college(M = 3.16, S.D. = 1.05). Medical workers report moderateto fairly high levels of job pressure, with a mean of 2.86(S.D. = .74) on a scale ranging from 1 (low) to 4 (high).Respondents have moderate to high job autonomy levels,with a mean of 3.05 (S.D. = .75) on a scale of 1 (low) to 4(high). The means for coworker support and supervisor

support are 3.48 (S.D. = .62) and 3.34 (S.D. = .67), respec-tively, on a scale from 1 (low) to 4 (high). Medical workersreport moderate work-to-life conflict on average, as themean is 2.52 (S.D. = .89) on a scale from 1 (low) to 5 (high).

Table 3OLS regression for the effects of workplace characteristics and work-to-life confli

Variables Model 1

B S EB

Work-to-life conflict – –

Job pressure .15 .05

Job autonomy .01 .05

Work hours −.001 .002 −Nonstandard hours −.08 .07 −Coworker support −.05 .06 −Supervisor support −.15 .06 −Age −.01 .003 −Race .18 .08

Gender −.08 .07 −Child under 18 −.09 .07 −Education −.05 .04 −Household income −.03 .06 −R2 .178

Change in R2 –

F for model 4.20***

F for change in R2 –

Source: The National Study of the Changing Workforce 2002.Note. Nonstandard hours is a dummy variable coded 1 if the respondent worksrespondent is white and 0 if the respondent is non-white. Gender is a dummy vaa dummy variable coded 1 if the respondent has a child under 18 living in the ho

* p < 0.05.** p < 0.01.

*** p < 0.001.

The majority of medical workers report low to moderatepsychological distress, with a mean of 2.18 (S.D. = .51) on ascale of 1 (low) to 5 (high). The bivariate correlations arepresented in Table 2.

5.2. Results of the regression analysis

To test the first two hypotheses, psychological distressis regressed on job pressure, work hours, nonstandard

work hours, job autonomy, coworker support, supervisorsupport, and the control variables (Model 1 in Table 3).Hypothesis 1, focusing on job demands, states that jobpressure, work hours, and nonstandard work hours are

ct on the psychological distress of medical workers (N = 246).

Model 2

B S EB ˇ

– .27 .04 .45***

.20** .03 .05 .04

.02 .02 .04 .03

.04 −.003 .002 −.07

.08 −.07 .06 −.05

.06 .03 .06 .03

.19** −.09 .05 −.11

.24*** −.01 .003 −.17**

.13* .14 .08 .10

.06 −.08 .07 −.06

.09 −.13 .06 −.13*

.09 −.05 .03 −.09

.04 −.07 .05 −.09.304.127

7.81***

42.17***

nonstandard hours. Race is a dummy variable that was coded 1 if theriable coded 0 if the respondent is female and 1 if male. Child under 18 isme and 0 if not.

l Scienc

pcao2onm2c

5

ttaitirmbaujaarves2flpprrctrcaitft2iSt2ssspHsdw

K.L. Minnotte et al. / The Socia

ositively related to psychological distress among medi-al workers. The results partially support this hypothesis,s there is a positive and significant relationship betweennly job pressure and psychological distress. Hypothesis, which focuses on job resources, states that job auton-my, coworker support, and supervisor support will beegatively associated with psychological distress amongedical workers. The results partially support Hypothesis

, as only one job resource—supervisor support—is signifi-antly associated with the dependent variable.

.3. Results of the mediation analysis

Model 2 shows relationships between work charac-eristics and psychological distress with the addition ofhe mediating variable of work-to-life conflict. There are

few requirements that are needed to establish mediat-ng relationships (Baron & Kenny, 1986). In establishinghese requirements, we focus on workplace character-stics that are significant in Model 1. First, significantelationships between the independent variables and theediating variable must be established, which is tested

y regressing work-to-life conflict on the workplace char-cteristics and the control variables (full results availablepon request). There is a positive association between

ob pressure and work-to-life conflict ( = .34, p < .001),nd a negative relationship between supervisor supportnd work-to-life conflict ( = −.18, p < .01). Thus, the firstequirement in establishing mediation is met for super-isor support and job pressure. The second condition instablishing mediation requires that relationships that areignificant in Model 1 become non-significant in Model. This requirement is met, as when work-to-life con-ict is added to the model, the relationships between jobressure and psychological distress and supervisor sup-ort and psychological distress are not significant. A thirdequirement in establishing mediation calls for a significantelationship between the mediating variable (work-to-lifeonflict) and the dependent variable (psychological dis-ress). This requirement is fulfilled because there is a directelationship between work-to-life conflict and psychologi-al distress among medical workers in Model 2. This findinglso supports Hypothesis 3 that posits work-to-life conflicts directly and positively associated with psychological dis-ress. The final requirement in establishing mediation callsor a significant increase in the explained variance whenhe mediating variable is added. The findings from Model

show a .127 unit increase in the R2, and this increases statistically significant. Thus, this requirement is met.obel tests were also undertaken to further establish work-o-life conflict as a mediating variable (Preacher & Hayes,004). The Sobel tests suggest that work-to-life conflict is aignificant mediator of the relationships between job pres-ure and psychological distress (Sobel = 4.47, p < .001) andupervisor support and psychological distress (Sobel = 2.60,

< .01). Altogether, the mediation findings suggest that

ypothesis 4 is partially supported and that the relation-

hips between job pressure and supervisor support and theependent variable psychological distress are mediated byork-to-life conflict.

e Journal 50 (2013) 408–417 415

6. Discussion and conclusion

The goal of this study is to examine the relationshipsbetween workplace characteristics, work-to-family con-flict, and psychological distress among workers in themedical industry. In particular, we address if workplacecharacteristics impact psychological distress by shapingwork-to-life conflict levels experienced by medical work-ers. Previous scholarship examines specific occupationalgroups in the medical industry, such as doctors and nurses(Geurts et al., 1999; Janssen et al., 2004; Montgomery et al.,2006). This study takes a different approach by lookingat workers in the broader medical industry to examinethe direct relationships between workplace characteris-tics and psychological distress and the potential mediatingrole of work-to-life conflict. In doing so, we contribute tothe literature by examining the industry in a broad man-ner, by incorporating key control variables, and by taking amore nuanced approach to examine workplace social sup-port.

The results of this study show direct relationshipsbetween job pressure, supervisor support, and psycholog-ical distress among medical workers, which is consistentwith past research that emphasizes these workplacecharacteristics’ importance (Diestel & Schmidt, 2009;Escribà-Agüir & Pérez-Hoyos, 2007). This study under-scores the salience of workplace social support, especiallysupervisor support, in predicting psychological distressamong workers in the medical industry. Supervisor supportis so crucial that past research suggests that the well-beingof employees and even their families may be impactedby their immediate supervisors, as supervisors are oftengatekeepers in workers’ ability to rearrange work sched-ules, thereby, either enhancing or detracting from workers’attempts to balance work and family responsibilities (Ryan& Kossek, 2008).

For the indirect relationships, our results suggest thatwork-to-life conflict mediates the relationships betweentwo workplace characteristics—job pressure and super-visor support—and psychological distress for medicalworkers. This finding reinforces role theory along withVoydanoff’s (2002) model of work-to-life conflict as amediator of relationships between workplace character-istics and individual outcomes, such as psychologicaldistress. Other scholars also point to work-to-life conflict’srole as a mediator between work characteristics and psy-chological distress among specific types of medical workers(Geurts et al., 1999; Janssen et al., 2004; Montgomery et al.,2006). This study contributes to the literature by demon-strating the mediating relationship among a broader groupof medical workers, thereby, illustrating that work-to-lifeconflict issues are likely evident among medical workers ingeneral. We also use a nuanced exploration of workplacesocial support that separately examines support stemmingfrom coworkers and supervisors, with our findings sug-gesting that only supervisor support is indirectly relatedto psychological distress through the mediating variable

work-to-life conflict. This is likely because supervisors havemore power to shape the actual workplace context, and thestressors inherent therein, than coworkers usually do, andhence supervisor support may be especially important to

l Scienc

416 K.L. Minnotte et al. / The Socia

those in the medical industry who work under such stress-ful conditions.

Altogether, the findings from this study highlight thatalleviating work-to-life conflict among medical workersis important. As such, appropriate action should be takento decrease such conflict in order to reduce issues of jobburnout and to increase quality of patient care. Both indi-viduals and organizations should make changes to copewith issues relating to work-to-life conflict and psycho-logical distress. One possibility for employers to consideris to move beyond simply implementing formal organiza-tional policies that support workers navigating work andpersonal life. In particular, it may be more important foremployers to work toward informally changing the normsand culture of the workplace, such that employees feelcomfortable using existing work–life policies (Thompson,Beauvais, & Lyness, 1999), leading to a workplace that isoverall more supportive of workers attending to personalresponsibilities. Along these lines, medical workers andmedical facilities both stand to benefit from policies thatpromote formal and informal workplace social support,and enhancing supervisor support should be emphasized.Perhaps more closely monitoring relationships betweensupervisors and employees may help ensure that theserelationships are supportive. This may be particularly help-ful if medical workers do not use available work–lifepolicies due to fear of informal reprimands from supervi-sors (Thompson et al., 1999). Providing supervisors witheducation about the stressors faced by their employees andthe role of work-to-life conflict in stress processes may alsobe beneficial. Additionally, medical facilities should imple-ment policies and practices that reduce job pressure, suchas hiring more people to carry out tasks that are often theresponsibility of one person. Hiring more employees tohelp out with patient care and administrative tasks mayhelp ease medical workers’ feelings of high pressure onthe job. Altogether, efforts aimed at reducing work-to-lifeconflict, reducing job pressure, and increasing supervisorsupport may help address the unique stressors experiencedby medical workers, which may ultimately enhance thequality of care they provide and increase job tenure.

A few limitations of this study should be noted. First,because the design of the study is cross-sectional, directionand causality cannot be established. Without longitudi-nal data we are unable to firmly determine that therelationships proposed do not work in the reverse, withthe dependent variable, psychological distress, leading tochanges in the independent variables, such as supervisorsupport, as it is plausible that workers exhibiting signsof such distress might prove burdensome for supervisors.Such supervisors, then, might reduce their support whenpsychological distress becomes evident among workers.We do note that most theoretical work in this area positsworkplace characteristics, including workplace social sup-port, as predictors of worker well-being rather thanoutcomes of worker well-being (Voydanoff, 2002, 2007).Second, because psychological distress is a global measure,

it is unclear what specific types of distress are experi-enced by workers, which may complicate the process ofimproving situations for these workers. Third, our studycontributes to the literature by looking at the overall

e Journal 50 (2013) 408–417

medical industry rather than specific occupations, how-ever, we have no way of knowing whether the occupationsof the respondents are representative of the distribu-tion of occupations in the overall medical industry. Someoccupations may be underrepresented in this study andothers may be overrepresented. Lastly, although quantita-tive research has many strengths, qualitative research, suchas conducting interviews and engaging in observation ofworkplace practices, would offer further insights into theunique work stressors experienced by medical workers.

This study underscores that continuing to examinework–life issues among medical workers is important.Future research would do well to further untangle theserelationships. Because conflict between work and per-sonal life is bi-directional, future research exploringantecedents and consequences of life-to-work conflict, inwhich demands from the personal life domain conflict withdemands from the work domain, is especially promising.Further, including other organizational (e.g., availability ofother “family-friendly” policies) and individual level (e.g.,personality characteristics) factors that influence psycho-logical distress among medical workers may provide amore holistic picture of the psychological distress expe-rienced in the medical industry.

References

Aiken, L. H., & Sloane, D. M. (1997). Effects of organizational innovationsin AIDS care on burnout among urban nurses. Work & Occupations, 24,453–477.

Akerstedt, T., Fredlund, P., Gillberg, M., & Jansson, B. (2002). Work loadand work hours in relation to disturbed sleep and fatigue in a largerepresentative sample. Journal of Psychosomatic Research, 53, 585–588.

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 aspecial emphasis on cross-domain versus matching-domain relations.Journal of Occupational Health Psychology, 16, 151–169.

Baron, R., & Kenny, D. A. (1986). The moderator-mediator variable dis-tinction in social psychological research: Conceptual, strategic, andstatistical considerations. Journal of Personality and Social Psychology,51, 1173–1182.

Beutell, N. J. (2010). Health, supervisory support, and workplace culturein relation to work–family conflict and synergy. Psychological Reports,107, 3–14.

Bildt, C., & Michelsen, H. (2002). Gender differences in the effects fromworking conditions on mental health: A 4-year follow-up. Interna-tional Archives of Occupational and Environmental Health, 75, 252–258.

Bourbonnais, R., Comeau, M., & Vézina, M. (1999). Job strain and evolu-tion of mental health among nurses. Journal of Occupational HealthPsychology, 4, 95–107.

Burke, R. J., & Greenglass, E. R. (1999). Work–family conflict, spouse sup-port, and nursing staff well-being during organizational restructuring.Journal of Occupational Health Psychology, 4, 327–336.

Craig de Silva, E., Clark, E. J., Whitaker, T., Wilson, M., Arrington, P., &Nkabyo, G. (2008). Stress at Work: How Do Social Workers Cope? NASWMembership Workforce Study,. Washington, DC: National Associationof Social Workers.

Dawson, S. (2012). Aging America creates demand for healthcare workers.http://www.nlm.nih.gov/medlineplus/news/fullstory 121477.html

Diestel, S., & Schmidt, K. H. (2009). Mediator and moderator effects ofdemands on self-control in the relationship between work load andindicators of job strain. Work and Stress, 23, 60–79.

Erickson, J., Martinengo, G., & Hill, E. (2010). Putting work and family expe-riences in context: Differences by family life stage. Human Relations,63, 955–979.

Escribà-Agüir, V., & Pérez-Hoyos, S. (2007). Psychological well-being and

psychosocial work environment characteristics among emergencyand medical nursing staff. Stress and Health, 23, 153–160.

Fenwick, R., & Tausig, M. (2001). Scheduling stress: Family and healthoutcomes of shift work and schedule control. American BehavioralScientist, 44, 1179–1198.

l Scienc

F

F

F

G

G

G

H

H

H

H

H

I

J

J

K

L

M

M

M

M

M

P

P

K.L. Minnotte et al. / The Socia

ielding, J., & Weaver, S. (1994). A comparison of hospital- andcommunity-based mental health nurses: Perceptions of their workenvironment and psychological health. Journal of Advanced Nursing,19, 1196–1204.

ord, G. A., III. (2013). Physician burnout: ‘The call’ derailed? SouthernMedical Journal, 106, 295–296.

ord, M. T., Heinen, B. A., & Langkamer, K. L. (2007). Work and familysatisfaction and conflict: A meta-analysis of cross-domain relations.Journal of Applied Psychology, 92, 57–80.

eurts, S., Rutte, C., & Peeters, M. (1999). Antecedents and consequencesof work–home interference among medical residents. Social Scienceand Medicine, 48, 1135–1148.

rant-Vallone, E. J., & Ensher, E. A. (2001). An examination of work andpersonal life conflict, organizational support, and employee healthamong international expatriates. International Journal of InterculturalRelations, 25, 261–278.

reenhaus, J. H., & Beutell, N. J. (1985). Sources of conflict between workand family roles. The Academy of Management Review, 10, 76–88.

aines, V., III, Marchand, A., Rousseau, V., & Demers, A. (2008). The medi-ating role of work-to-family conflict in the relationship betweenshiftwork and depression. Work and Stress, 22, 341–356.

ämmig, O., Gutzwiller, F., & Bauer, G. (2009). Work–life conflict and asso-ciations with work-and nonwork-related factors and with physicaland mental health outcomes: A nationally representative cross-sectional study in Switzerland. BMC Public Health, 9, 435–450.

ill, E. J. (2005). Work–family facilitation and conflict, working fathers andmothers, work–family stressors and support. Journal of Family Issues,26, 793–819.

ilton, M. F., Whiteford, H. A., Sheridan, J. S., Clearly, C. M., Chant, D. C.,Wang, P. C., & Kessler, R. C. (2008). The prevalence of psychologicaldistress in employees and associated occupational risk factors. Journalof Occupational & Environmental Medicine, 50, 746–757.

ughes, E. L., & Parkes, K. R. (2007). Work hours and well-being: The rolesof work-time control and work–family interface. Work & Stress, 21,264–278.

lies, R., & Dimotakis, N. (2010). Psychological and physiological reactionsto high workloads: Implications for well-being. Personnel Psychology,63, 407–436.

anssen, P. P. M., Peeters, M. C. W., de Jonge, J., Houkes, I., & Tum-mers, G. E. R. (2004). Specific relationships between job demands,job resources, and psychological outcomes and the mediating role ofnegative work–home interference. Journal of Vocational Behavior, 65,411–429.

ohnson, C.K. (2010). Facing doctor shortage: 28 states may expand nurse’srole. http://www.aanp.org/NR/rdonlyres/C9472AAF-13B6-4CD3-BC98-2DFB053CF83E/3950/4 16 10USATodaycomArticle.pdf

innunen, U., Geurts, S., & Mauno, S. (2004). Work-to-family conflict andits relationship with satisfaction and well-being: A one-year longitu-dinal study on gender differences. Work and Stress, 18, 1–22.

inzer, M., Gerrity, M., Douglas, J. A., McMurray, J. E., Williams, E. S., & Kon-rad, T. R. (2002). Physician stress: Results from the physician worklifestudy. Stress and Health, 18, 37–42.

ajor, V. S., Klein, K. J., & Ehrhart, M. G. (2002). Work time, work inter-ference with family, and psychological distress. Journal of AppliedPsychology, 87, 427–436.

innotte, K. L. (2012a). Perceived discrimination and work-to-life conflictamong workers in the U.S. Sociological Quarterly, 53, 188–210.

innotte, K. L. (2012b). Family structure and the work–family interface:Work-to-family conflict among single and partnered parents. Journalof Family and Economic Issues, 33, 95–107.

irowsky, J., & Ross, C. E. (1995). Sex differences in distress: Real or arti-fact? American Sociological Review, 60, 449–468.

ontgomery, A. J., Panagopolou, E., & Benos, A. (2006). Work–family inter-ference as a mediator between job demands and job burnout amongdoctors. Stress and Health, 22, 203–212.

arasuraman, S., Purohit, Y. S., Godshalk, V. M., & Beutell, N. J. (1996). Work

and family variables, entrepreneurial career success, and psychologi-cal well-being. Journal of Vocational Behavior, 48, 275–300.

arasuraman, S., & Simmers, C. A. (2001). Type of employment,work–family conflict and well-being: A comparative study. Journal ofOrganizational Behavior, 22, 551–568.

e Journal 50 (2013) 408–417 417

Parikh, P., Taukari, A., & Bhattacharya, T. (2004). Occupational stressand coping among nurses. Journal of Health Management, 6,115–127.

Pisarski, A., Brook, C., Bohle, P., Gallois, C., Waston, B., & Winch, S. (2006).Extending a model of shift-work tolerance. Chronobiology Interna-tional, 23, 1363–1377.

Pomaki, G., Supeli, A., & Verhoeven, C. (2007). Role conflict and healthbehaviors: Moderating effects on psychological distress and somaticcomplaints. Psychology and Health, 22, 317–335.

Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for esti-mating indirect effects in simple mediation models. Behavior ResearchMethods, Instruments, & Computers, 36, 717–731.

Presser, H. B. (2003). Working in a 24/7 economy: Challenges for Americanfamilies. New York: Russell Sage Foundation.

Qidwai, W., Beasley, J. W., & Gomez-Clavelina, F. J. (2008). The presentstatus and future role of family doctors: A perspective from theInternational Federation of Primary Care Research Networks. PrimaryHealth Care Research & Development, 9, 172–182.

Rees, D. W. (1995). Work-related stress in health service employees. Jour-nal of Managerial Psychology, 10, 4–11.

Roopalekha, P. N., Latha, K. S., & Swetha, P. (2012). Occupational stress andcoping among nurses in a super specialty hospital. Journal of HealthManagement, 14, 467–479.

Rout, U. (2000). Stress amongst district nurses: A preliminary investiga-tion. Journal of Clinical Nursing, 9, 303–309.

Ryan, A. M., & Kossek, E. E. (2008). Work–life policy implementation:Breaking down or creating barriers to inclusiveness. Human ResourceManagement, 47, 295–310.

Schieman, S., & Glavin, P. (2011). Education and work–family conflict:Explanations, contingencies, and mental health consequences. SocialForces, 89, 1341–1362.

Shields, M. (2002). Shift work and health. Health Reports, 13, 11–33.Shields, M., & Ward, M. (2001). Improving nurse retention in the national

health service in England: The impact of job satisfaction on intentionsto quit. Journal of Health Economics, 20, 677–701.

Sparks, K., Cooper, C., Fried, Y., & Shirom, A. (1997). The effects of hoursof work on health: A meta-analytic review. Journal of Occupational &Organizational Psychology, 70, 391–408.

ter Doest, L., & de Jonge, J. (2006). Testing causal models of job characteris-tics and employee well-being: A replication study using cross-laggedstructural equation modeling. Journal of Occupational and Organiza-tional Psychology, 79, 499–507.

Thomas, L. T., & Ganster, D. C. (1995). Impact of family-supportive workvariables on work–family conflict and strain: A control perspective.Journal of Applied Psychology, 80, 6–15.

Thompson, C., Beauvais, L. L., & Lyness, K. S. (1999). When work–familybenefits are not enough: The influence of work–family culture on ben-efit utilization, organizational attachment, and work–family conflict.Journal of Vocational Behavior, 54, 392–415.

Thompson, C. A., & Prottas, D. J. (2005). Relationships among orga-nizational family support, job autonomy, perceived control, andemployee well-being. Journal of Occupational Health Psychology, 10,100–118.

Tiedje, L. B., Wortman, C. B., Downey, G., Emmons, C., Biernat, M., & Lang, E.(1990). Women with multiple roles: Role-compatibility perceptions,satisfaction, and mental health. Journal of Marriage and Family, 52,63–72.

Voydanoff, P. (2002). Linkages between the work–family interface andwork, family, and individual outcomes. Journal of Family Issues, 23,138–164.

Voydanoff, P. (2005). Consequences of boundary-spanning demands andresources for work-to-family conflict and perceived stress. Journal ofOccupational Health Psychology, 10, 491–503.

Voydanoff, P. (2007). Work, family, and community: Exploring interconnec-tions. Mahwah, NJ: Lawrence Erlbaum Associates.

Wilson, M. G., DeJoy, D. M., Vandenberg, R. J., Richardson, H. A., &

McGrath, A. L. (2004). Work characteristics and employee health andwell-being: Test of a model of health work organization. Journal ofOccupational Psychology, 77, 565–588.

Woods, R. (2009). Industry output and employment projections to 2018.Monthly Labor Review, 132, 52–81.