job stress among female flight attendants

Job Stress Among Female Flight Attendants

Leslie A. MacDonald, ScDJames A. Deddens, PhDBarbara A. Grajewski, PhDElizabeth A. Whelan, PhDJoseph J. Hurrell, PhD

We evaluated the presence of chronic job stressors among flightattendants (FAs) to examine the relationships between these job stressorsand psychological distress and job dissatisfaction. Seventy-three femaleFAs (90% participation) employed at two commercial airlines completeda detailed questionnaire. Standard questions and scale measures wereused to assess job stressors, psychological distress, and job dissatisfac-tion. The association between job stressors and these outcomes wasevaluated using multiple regression analysis. Except for fatigue, distressand job dissatisfaction were moderate to low. Job stressors were found tohave a substantive effect on these outcomes, following adjustment forindividual factors. Despite moderate-to-low levels of distress and dissat-isfaction, targeted efforts to reduce selected job stressors and to enhancesocial support may be important steps toward improving the well-beingand satisfaction of FAs. (J Occup Environ Med. 2003;45:703–714)

T here is a paucity of information inthe scientific literature concerningflight attendants’ exposure to jobstressors and psychological distressoutcomes. Although the flying publicregularly observes flight attendantsperforming activities associated withroutine passenger service, the criticalpublic safety role of flight attendantsand the concomitant demands oftengo unrecognized. These demands in-clude unobtrusive and highly disci-plined responses to medical andother emergencies, vigilance for ac-tivities within the cabin environmentthat may accidentally or deliberatelythreaten the safety of passengers orthe flight crew, assurance of passen-ger compliance with Federal aviationregulations, and responses to passen-ger “air rage.”

In the few published investigationsof flight attendants’ working condi-tions, more overt stressors, such ascritical flight incidents and physicaljob stressors (noise, vibration, lim-ited working space), are most oftenhighlighted. For example, anxiety,flight phobias, and post-traumaticstress have been reported among aircrew members (including flight at-tendants) after flight incidents.1,2

Even in the absence of such inci-dents, up to one third (37%) of flightattendants often feel anxious beforetake-off.3 Suvanto and Ilmarinen re-port that cognitive as well as physi-cal job demands are a source ofstress among flight attendants, notingdemands for foreign language skills,the need to make decisions rapidly,and the importance of general educa-tion and professional experience inwork execution.4 A study conductedin Norway showed that only half ofthe flight attendants surveyed weresatisfied with supervision and social

From the National Institute for Occupational Safety and Health, Division of Surveillance, HazardEvaluations and Field Studies, Industrywide Studies Branch (Dr MacDonald, Dr. Deddens, DrGrajewski, Dr Whelan, Dr Hurrell), and University of Cincinnati, Department of MathematicalSciences (Dr Deddens), Cincinnati, Ohio.

This article was written by an officer or employee of the U.S. Government as part of his/her officialduties and is therefore not subject to U.S. copyright.

Address correspondence to: Leslie MacDonald, ScD, Research Ergonomist, Industrywide StudiesBranch, Division of Surveillance, Hazard Evaluations and Field Studies, 4676 Columbia Parkway, MSR-15, Cincinnati, OH 45226-1998; E-mail: [email protected].

DOI: 10.1097/01.jom.0000071509.96740.dd

JOEM • Volume 45, Number 7, July 2003 703

support, and dissatisfaction was es-pecially high concerning the sharingof information related to specific op-erational procedures and technicalmatters.5 Findings from a recentstudy from Italy suggested that ex-tended absences from home havedeleterious effects on the close per-sonal relationships of flight atten-dants and that flight attendants oftenreport feelings of isolation and lone-liness.6

Because previous studies havebeen limited to examinations of thepsychological reaction to criticalflight events or to descriptive reportsof job stress conditions, this studysought to evaluate the presence ofchronic job stressors among activelyemployed female flight attendants inthe United States and to examine therelationships between these job stres-sors and psychological distress andjob dissatisfaction. It was hypothe-sized that job stressors would bepresent, that these stressors would beassociated with psychological dis-tress and job dissatisfaction even af-ter adjustment for individual factors,and that selected factors may miti-gate the putative effects of job stres-sors. The specific goals of this studywere: (1) to identify specific jobstressors in flight attendant work forinclusion in larger reproductive epi-demiology studies among this occu-pational group; and (2) to examinewhether these job stressors predictpsychological distress, perceivedstress, and job dissatisfaction afteradjustment for individual factors.This investigation is part of a seriesof studies by National Institute forOccupational Safety and Health(NIOSH; Cincinnati, OH) to charac-terize flight attendants’ exposure tofactors that may affect reproductiveand other health endpoints.

Data for this study were collectedbefore the events of September 11,2001, or “9/11,” when terrorists hi-jacked four commercial domesticflights, claiming the lives of all flightcrew and passengers and more than2500 people who occupied targetedbuildings. The events of 9/11 under-

score the real potential for flightattendants to be directly exposed toviolence in the workplace. More-over, the aftermath of 9/11 is char-acterized by an environment of in-creased threats to flight crew andpassenger safety, a decline in airtravel, and ensuing mass layoffs ofairline personnel, airline bankrupt-cies, and a proposal to permit armedpilots on board commercial passen-ger planes. These conditions marksignificant shifts in the work envi-ronment of flight attendants not ad-dressed in this study that could haveimportant implications for theirhealth and welfare.

Methods

Study ParticipantsStudy data were obtained from 73

female flight attendants employed attwo commercial airlines in theUnited States. Signed informed con-sent was obtained from each partici-pant. Forty-five of the flight atten-dants (62%) were enrolled from twoairlines in a biological monitoringfeasibility study of menstrual func-tion (study A) during 1995, with dataobtained by self-administered ques-tionnaire. As described in detail byWhalen et al.,7 participants enrolledinto study A were randomly selectedfrom employer lists that were gener-ated according to demographic char-acteristics (aged 18 to 45 and fe-male), employment characteristics(full-time flight attendant), and flightpatterns maximizing the range of twofactors of interest for the menstrualfunction study, cosmic radiation, andtravel through multiple time zones.Additional selection criteria includedpregnancy status (not pregnant), cur-rently having menstrual cycles, andnot using oral contraceptives or anintrauterine device. Eligible womenwere enrolled until the target samplesize was met.

The remaining study participants(n � 28, 38%) were enrolled fromone of the study A airlines in afeasibility study of exposure to jobstress conditions and physical work

demands (study B), with data on thesame battery of measures obtainedby telephone interview. Participantsin study B were all flight attendantsselected from the employer’s flightroster for one of five specific round-trip flights in 1995 or 1996 chosenaccording to flight duration andflight route (ie, domestic or interna-tional). Seven males who partici-pated in study B were excluded fromthis analysis as a result of smallnumbers. Participation rates amongeligible flight attendants for the twostudies were 42% and 90%, respec-tively. The lower participation ratefor study A was likely the result ofsubstantial study demands involvingthe collection and storage of dailybiologic samples and daily diary re-cordings for at least one month, asdescribed further elsewhere.7

Study MeasuresThe study questionnaire included

over 50 measures. Nonoccupationalmeasures included individual factors(age, race, education, height, weight,coping style), nonwork stressors (re-cent major life events, conflict orhostility in the home, preschool-agedchildren in the home), and level ofsocial support outside of the work-place. The scale measures for copingstyle (active and passive) and sup-port outside of work were selectedfrom the NIOSH Generic Job StressQuestionnaire (NGJSQ).8 Partici-pants were also asked to report thenumber of block hours (taxi andairborne time) that they typicallyworked per month, whether theyworked primarily on domestic or in-ternational routes and, in study B, thedate they started employment as aflight attendant.

Job StressorsJob stressors, which were treated

as independent variables in thisstudy, were assessed primarily fromresponses to standard questions ormulti-item scales (Appendix I). Scalemeasures of role ambiguity, role con-flict, supervisor support, coworkersupport, and task control were se-

704 Job Stress Among Female Flight Attendants • MacDonald et al

lected from the NGJSQ. Scale mea-sures of mental demands, psycholog-ical job demands, decision latitudeand job strain (ratio of psychologicaljob demands and decision latitude)were selected from the Job ContentQuestionnaire.9 A short version ofthe emotional load and the emotionalrewards scale measures were se-lected from The Emotions at WorkScales by Spratt.10 Internal reliabilityof the scale measures was evaluatedfor use in this study population usingCronbach’s standardized � coeffi-cient. The � coefficient for four scalemeasures (role ambiguity, role con-flict, psychological job demands,emotional load) were judged to beunacceptably low (�0.61). For eachof these four constructs, single-itemmeasures from each multi-item scalewere selected for use in statisticalanalyses based on the magnitude ofthe correlation of the items with thefull-scale measure.

In addition to the scale measures, asingle-item measure of hostility (Iam exposed to conflict or hostilityfrom the people I work with) wasadapted from the 1993 NorthwesternNational Life Insurance CompanyFear and Violence in the WorkplaceSurvey (Minneapolis, MN; Appen-dix I). A composite measure of im-balance between work and nonworkobligations was developed from twoitems (the demands of my job createserious stress in my family and con-flicts exist between job demands andoutside obligations), which wereadapted from Pilch.11 A compositemeasure of job insecurity was devel-oped from two items (my job secu-rity is good, reverse scored) and(during the past year, did you facejob loss or layoff?), which were se-lected from the NGJSQ.

In addition to the standard ques-tions and scales, information relatedto unusual or transient events wasobtained by daily diary among par-ticipants of study A during the 1- to3-month study period. During work-days, the daily diary obtained flight-specific schedule information, andflight attendants were asked “did

anything unusual happen on today’sflight(s) (flight problems, injuryetc.)?” Those with an affirmativeresponse were asked to provide anexplanation. Participants of study Bwere asked an open-ended question,which read as follows: “Are therepsychological or emotional stressesin your job that we have not askedyou about that you feel may put yourhealth or safety at risk?” Events oractivities listed in the daily diary andresponses to the open-ended questionserved to identify job stressor condi-tions not previously identified and toimprove the context-specific inter-pretation of the quantitative ques-tionnaire-based measures.

Psychological Distress,Perceived Stress, and JobDissatisfaction

Measures of psychological dis-tress, perceived stress and job dissat-isfaction were treated as dependentvariables in this study. Multi-itemadjective scales were selected fromthe Profile of Mood States (POMS)for the assessment of psychologicaldistress (Appendix II). Three ques-tions were selected for each distressscale based on factor loadings ofprevious studies (POMS Users Man-ual, 1981, Educational and IndustrialTesting Service, San Diego, CA). Ashort version of the Perceived StressScale was selected to obtain a globalmeasure of the degree to which par-ticipants appraised life situations inthe recent past as stressful.12 Theinternal consistency of these scalemeasures was evaluated for use inthis study population using Cron-bach’s standardized � coefficient. Aone-item measure of job satisfactionwas selected from the NGJSQ.

Data AnalysisBefore combining the data from

the two feasibility studies, homoge-neity of the data was assessed bycomparing the means of the individ-ual factors, nonwork stressors, work-related factors (eg, block hours,

flight routes), job stressors and de-pendent variables using t tests andWilcoxon nonparametric tests and bycomparing the slope of the relationbetween each independent and de-pendent variable using “study” as aninteraction term. Although two inde-pendent variables (mental demandsand supervisor support) and one de-pendent variable (fatigue) had signif-icantly different means between thetwo study populations, there was nostatistically significant difference inthe relation (slope) between the in-dependent and dependent variablesin the two study populations. Allvariables were retained for subse-quent analyses.

Visual inspection of the residualswas performed for all dependentvariables and depression was judgedto be highly skewed. After a log10

transformation was applied to thedepression scale, the residuals werefound to conform with the normalityassumption of linear regression anal-ysis.13 Pearson correlation coeffi-cients (r) were computed for theassessment of covariation among allindependent and dependent vari-ables.

Multiple linear regression analysiswas used to examine the multivariaterelation between job exposure fac-tors and psychological distress, per-ceived stress and job dissatisfaction,adjusting for all individual factorsand nonwork stressors showing atleast a moderate association (P value� 0.20). Variables were retained inthe final models at P value �0.05.Multiple regression analyses also in-cluded an exploratory assessment ofthe role of factors such as copingstyle, social support, emotional re-wards and job control in mitigatingthe putative effects of adverse jobconditions. Due to small numbers,these models were comprised of in-teraction terms and lower-order com-ponents only. Interaction terms wereretained at p � 0.10. Data analyseswere performed using PC SAS soft-ware, version 8.1 (SAS, Cary, NC).


Results

Study ParticipantsData on participant characteristics

(individual factors, nonwork stres-sors, and employment) are summa-rized in Table 1. The average age ofthe study participants was 37 years,ranging from 27 to 55 years. Averagebody mass index (BMI) or estimateof body fat based on height andweight was 21 kg/m2, and rangedfrom 18 to 26 (a healthy BMI foradults is between 18.5 and 24.9 kg/m2, based on the effect of bodyweight on disease).14 Although mostparticipants adopted some degree ofboth active and passive coping strat-

egies when faced with problems, ac-tive coping was a more prevalentcoping style. Social support outsidework was high, and the presence ofconflict at home was moderately low(about one third of participants). Par-ticipants worked an average of 80block hours per month and half(49%) worked primarily on domesticflight routes. The majority (92%) ofthe participants were Caucasian andone third had a college degree. Ap-proximately one third (36%) hadchild dependents (any age) living athome and 11% had at least onepreschool-aged child. Recent majorlife change (change in marital status,serious illness or loss of friend orfamily member) that was at least

“somewhat stressful” was reportedby 44% of the participants.

Job StressorsResults from the assessment of job

stressors are summarized in Table 2.Role ambiguity (there are clearplanned goals and objectives for myjob, reverse scored) and role conflict(receive incompatible requests, inad-equate resources, rules or policiesbent to complete assignment) werefairly low. Supervisor support wasmoderate, whereas coworker supportwas high. Mental demands (intenseconcentration required, tasks ofteninterrupted requiring attention at alater time, waiting for others slowsme down) were moderately high.Compared with normative data forworking women sampled in the USDepartment of Labor’s Quality ofEmployment Surveys, mean scoresfor psychological job demands weresignificantly higher among this sam-ple of flight attendants, whereasmean scores for decision latitude(job control) were not significantlydifferent.9 Emotional load (I have toseem concerned. . .) was high, as wasemotional rewards (when customerssay “thank you” it makes the jobworthwhile, working with people issatisfying). Nearly one third (29%)of the participants reported imbal-ance between their job demands andobligations outside of work (includ-ing family), and imbalance wasfound to occur more often amongthose with preschool-aged childrenin the home (63% vs. 25%, P �0.03).

Exposure CovariationThe magnitude of the correlation

between some job stressors showedevidence of exposure covariation orthe joint occurrence of job stressors(Table 3). Coincident exposure wasshown between mental demands andseveral other stressors: role conflict,low supervisor support, psychologi-cal job demands, decision latitudeand conflict at work. Low supervisorsupport was also related to role con-flict and emotional load. Imbalance

TABLE 1Study Participant Characteristics, n � 73

Mean SD Range Proportion

Age, years 37.1 6.6 27–55Body mass index (kg/m2) 21.1 1.6 18–26Active coping 3.5 1.0 1.5–5Passive coping 2.6 0.9 1–4.7Support outside work 10.7 1.6 6–12Conflict at home 1.7 0.8 1–4Block hours worked/month 80.1 14.9 40–115Job seniority, years* 13.5 9.3 2–34Fly primarily domestic routes 49%Caucasian 92%College degree 34%Pre-school aged child 11%Recent major life change 44%

* Data available only for participants of Study B, n � 28.

TABLE 2Descriptive Summary of Job Stressors, n � 73

Mean SD Range Proportion

Role ambiguity 2.2 1.5 1–7Role conflict 2.9 1.9 1–7Supervisor support 7.5 1.9 3–11Co-worker support 9.5 1.3 6–12Mental demands 7.2 1.9 4–12Psychological job demands 3.1 0.8 1–5Decision latitude 58.6 10.1 28–84Job strain 0.60 0.17 0.35–1.43Task control 14.2 4.1 5–25Emotional load 4.1 1.2 1–5Emotional rewards 4.1 0.8 2–5Conflict at work 2.2 0.75 1–4Imbalance 29%Job insecurity 27%


was inversely related to coworkersupport.

Level of Psychological Distress,Perceived Stress, and JobDissatisfaction

Except for fatigue, psychologicaldistress levels were found to be gen-erally low (Table 4). Fatigue wasmoderately high, followed by mod-erately low levels of anxiety andperceived stress. Depression scoreswere low. There were no reports offlight attendants being “not at allsatisfied” with their job, whereas38% reported being “very satisfied.”The correlations among the distressand perceived stress variables werehigh.

Bivariate AssociationsBivariate associations between the

independent variables (ie, individual

factors, nonwork stressors, work-related factors and job stressors) andpsychological distress, perceivedstress, and job dissatisfaction aresummarized in Table 5. Three jobstressors associated with all of thedistress measures were low supervi-sor support, mental demands, andimbalance. However, the associa-tions among the other independentvariables were more outcome spe-cific, such as between BMI and anx-iety and between low support outsideof work or job strain and depression.BMI, imbalance, and social relationsvariables (eg, low support outsidework, low coworker support, conflictat home) were associated with per-ceived stress, whereas factors associ-ated with job dissatisfaction includedrole ambiguity, low supervisor sup-port, low job control, low emotionalrewards, job strain, job insecurity,

and the absence of preschool-agedchildren in the home.

Multiple Regression ResultsFinal model results are shown in

Table 6. The amount of variance inthe distress measures explained bythe models ranged from 18% forfatigue to 39% for anxiety. The pro-portion of the variance explained forperceived stress and job dissatisfac-tion was 33% and 47% respectively.After accounting for individual fac-tors and non-work stressors, jobstressors that were important predic-tors of psychological distress in-cluded mental demands, psychologi-cal job demands, imbalance, and lowsupervisor support. Model entry re-strictions yielded a slightly differentset of predictors for anger and fa-tigue. Although the model entry re-strictions did not yield different re-sults for the other outcomes, it isimportant to note the coincident oc-currence of some job stressors in theinterpretation of these multiple re-gression findings.

Mediated AssociationsAn exploratory assessment of fac-

tors that may mediate the associa-tions reported above provide furtherevidence of the important role ofsocial relations to the occurrence ofthe outcomes of interest in this study.

TABLE 3Correlation Among Job Stressors, n � 73

Job Stressors 1 2 3 4 5 6 7 8 9 10 11 12 13 14

1. Role ambiguity –2. Role conflict 0.06 –3. Supervisor support �0.08 �0.39b –4. Co-worker support �0.24d 0.04 0.13 –5. Mental demands �0.21 0.27d �0.28d 0.05 –6. Psychological job

demands0.05 0.04 �0.02 0.03 0.28d –

7. Decision latitude �0.28d �0.02 0.23 0.18 0.30c 0.34 –8. Job strain 0.19 0.01 �0.18 �0.03 0.08 0.36c �0.75a –9. Task control �0.16 �0.02 0.13 0.15 �0.04 �0.09 0.49a �0.47a –

10. Emotional load �0.20 �0.21 0.28d 0.09 �0.23 �0.06 0.03 0.05 �0.19 –11. Emotional rewards �0.12 0.00 0.21 0.06 0.01 0.18 0.26d �0.08 0.11 0.01 –12. Conflict at work �0.20 0.17 0.02 �0.12 0.32c 0.27d 0.27d 0.17 0.18 0.14 �0.02 –13. Imbalance 0.00 0.19 �0.15 �0.34c 0.15 0.08 �0.11 0.08 �0.15 �0.01 �0.18 0.17 –14. Job Insecurity �0.04 �0.16 �0.27d �0.02 �0.04 �0.03 �0.16 0.24d �0.14 0.11 0.00 0.08 �0.09 –

P value: a �0.0001, b �0.001, c �0.01, d �0.05 (bold if P � 0.01).

TABLE 4Descriptive Summary of Dependent Variable Scores and Their Correlation,n � 73

Dependent Variables Mean 90th Percentile Range

Correlation

1 2 3 4 5 6

1. Anger 1.9 4 0–9 –2. Anxiety 2.5 5 0–9 0.55a –3. Fatigue 4.1 9 0–9 0.39b 0.63a –4. Depression 1.1 2 0–9 0.70a 0.60a 0.44a –5. Perceived stress 2.7 6 0–11 0.52a 0.67a 0.41b 0.63a –6. Job dissatisfaction 1.7 2 1–3 0.28d 0.23 0.27d 0.32c 0.15 –

P value: a � 0.0001, b � 0.001, c � 0.01, d � 0.05 (bold if p � 0.01).


Higher levels of support in relationsoutside the workplace was found toreduce the adverse effect of imbal-ance on anger, whereas lower levelsof support from coworkers and/orfrom relations outside the workplaceincreased the adverse effects oflower supervisor support on fatigueand job dissatisfaction. Lower levelsof supervisor support increased theadverse effect of imbalance and con-flict in relations at home on per-ceived stress, but higher task controldampened the adverse effect oflower levels of supervisor supportand job dissatisfaction.

Comparison of Domestic VersusInternational Flight Attendants

Half (51%) of the study partici-pants reported that they primarilyworked on international routes. Thenumber of self-reported blockhours worked per month washigher among flight attendants as-signed primarily to internationalroutes, compared with those pri-marily assigned to domestic routes

(84 hours vs. 76 hours). Interna-tional flight attendants were morelikely to be non-Caucasian, wereslightly older (P � 0.13), andslightly more likely to have a col-lege degree (P � 0.14), while do-mestic flight attendants wereslightly more likely to have a pre-school-aged child in the home (P �0.15). Job tenure, which was onlyavailable for participants of studyB (n � 28), was higher for flightattendants assigned to internationalroutes (mean � 15.3 years vs. 11.9years, P � 0.34). Job strain wassignificantly higher among domes-tic flight attendants, who hadhigher psychological job demands(P � 0.08) and somewhat lower jobcontrol (P � 0.27), but who werealso found to experience signifi-cantly higher levels of emotional re-wards from their work. Fatigue wassignificantly higher among the do-mestic flight attendants (P � 0.04),but no other differences in outcomeswere observed between domestic andinternational flight attendants. Be-

cause of small numbers of partici-pants, stratified multiple regressionanalyses were not performed.

Flight Attendants’ Descriptionsof Job Stress Conditions

Although the quantitative resultsdescribed above provide importantinformation to characterize job stressconditions, further insight into thespecific manner in which selectedjob stressors manifest within flightattendant work is provided by flightattendants’ daily diary reports of un-usual flight events (study A) andfrom the open-ended question “arethere psychological and emotionalstresses in your job that we have notasked you about that you feel mayput your health or safety at risk?(study B).”

Forty-two percent (19/45) of theparticipants from study A reportedthat an unusual event occurred on atleast one flight during the 1- to3-month study period. The mostcommon events reported were flightdelays and passenger-related needs

TABLE 5Bivariate Associations Between the Independent Variables and the Dependent Variables

Anger Anxiety Fatigue DepressionPerceived

StressJob

Dissatisfaction

Age 0.10 �0.06 �0.04 0.00 0.03 0.03Body mass index 0.11 0.33c 0.08 0.17 0.25d 0.02Education �0.01 0.00 �0.18 0.15 0.01 0.23Major life change 0.07 �0.10 �0.15 0.14 �0.02 0.05Pre-school child �0.14 0.08 �0.06 �0.21 0.03 �0.26d

Active coping �0.02 �0.14 0.05 �0.05 �0.17 0.03Passive coping 0.04 �0.09 �0.10 �0.09 �0.18 0.01Support outside work �0.19 �0.16 �0.06 �0.32c �0.32c �0.16Conflict at home 0.22 0.25d 0.12 0.19d 0.44a 0.08Block hours per month �0.01 �0.07 �0.04 0.01 �0.06 0.02Role ambiguity �0.05 0.06 0.10 0.10 0.02 0.24d

Role conflict 0.18 0.14 0.09 0.12 0.15 0.09Supervisor support �0.34c �0.34c �0.37b �0.34c �0.20 �0.51a

Co-worker support �0.07 �0.18 �0.02 �0.14 �0.24d �0.18Mental demands 0.50a 0.27d 0.36b 0.31c 0.19 0.14Psych. job demands 0.21 0.11 0.22 0.20 0.22 �0.17Decision latitude �0.04 �0.07 �0.16 �0.25d �0.15 �0.34c

Job strain 0.22 0.17 0.27d 0.35c 0.21 0.32c

Task control �0.18 �0.12 �0.21 �0.24d �0.13 �0.28d

Emotional load �0.01 �0.11 �0.21 �0.13 �0.15 0.22Emotional rewards �0.03 �0.14 �0.15 �0.06 �0.21 �0.31c

Conflict at work 0.15 0.20 0.17 0.00 0.16 �0.07Imbalance 0.31c 0.51a 0.28d 0.25d 0.46a 0.09Job insecurity 0.13 0.11 0.06 0.13 0.07 0.33c

P value: a �0.0001, b �0.001, c �0.01, d �0.05 (bold if P �0.01).


or problems. Flight delays were oftenreported to result from adverseweather conditions, but delays werealso reported to occur as the result ofequipment problems or difficult pas-senger experiences (eg, intoxica-tion). Many reports of passenger-related events occurred inconjunction with flight delays (eg,passenger deplaning, passenger an-ger over canceled flights, passengeranxiety about missed flight connec-tions). Other passenger-relatedevents included uncooperative be-havior (eg, resisting seat belt usewhen required, sitting in a non-assigned seat, intoxication, smokingin lavatory) and abusive behavior(eg, yelling and arguing with flightattendant, and one reported incidentof a flight attendant being spit on).One in-flight mechanical failure(blew a tire upon landing) and twoin-flight medical incidents (passen-

ger had a heart attack, passenger hiton the head with a briefcase) werereported. The crash of a commercialplane was reported by one flightattendant (a pilot she was flying withknew a crew member who waskilled).

Flight attendants’ responses to theopen-ended question included con-cerns about the threat of terrorism,infectious diseases (colds, tuberculo-sis), and weather or mechanicalemergencies in-flight. In addition,participants reported concern aboutpotential long-term health conse-quences of exposure conditions inthe cabin environment relating to airquality (eg, recirculating versus re-placement air, second-hand smoke),and cabin air pressure. Flight atten-dants also reported the direct impactsof flight schedules and travel-relateddemands on biological and socialrhythms. For example, flight atten-

dants reported that east-west flights“wear you out,” that sleep deficitscarry over into nonwork time, andthat travel logistics effectivelylengthen the workday (eg, difficul-ties with public transportation, locat-ing hotels). Absences associated withwork travel were reported to hinderthe development of nurturing rela-tionships, and reserve status was spe-cifically reported to complicate plan-ning for important social events.Conditions of social isolation anddetachment were also reported (eg,waking up in a “strange city,” lone-liness, and “sterile hotels”). Emo-tional load was reflected in the reportof specific passenger demands andresponsibilities (eg, unaccompaniedminors, cultural and language differ-ences). Crewmembers not workingtogether during emergencies, lack ofopen communication with the cock-pit, and performance reviews empha-sizing negative aspects of perfor-mance were examples ofunsupportive social relations cited byflight attendants.

DiscussionResults of this study showed that

levels of fatigue among actively em-ployed flight attendants were moder-ately high, whereas anger, anxiety,depression, perceived stress and jobdissatisfaction were moderately lowor low. Even for the less prevalentoutcomes, a considerable portion oftheir variance was predicted by mod-ifiable job stressors: mental and/orpsychological job demands, imbal-ance between job demands and out-side obligations, low supervisor sup-port, and role ambiguity andemotional load (job dissatisfaction).Several of these job stressors werefound to covary, thus it may be thejoint effect of these stressors that isimportant for the prediction of dis-tress. The presence of social support(from all measured sources) wasfound to reduce the adverse effect ofsome job stressors while the absenceof support was found to exacerbatethe effect of other stressors. Al-though psychological distress and

TABLE 6Multiple Regression Modeling of Psychological Distress, Perceived Stress, andJob Dissatisfaction

ModelSemipartial

R2Standard

ErrorP

Value

Anger (R2 � 0.30)*Mental demands 0.38 0.25 0.08 �0.01Imbalance 0.91 0.05 0.43 0.04

Anxiety (R2 � 0.39)Body mass index 0.32 0.11 0.13 0.02Imbalance 2.02 0.21 0.47 �0.01Supervisor support �0.31 0.07 0.11 0.01

Fatigue (R2 � 0.18)†Supervisor support �0.36 0.13 0.16 �0.01Mental demands 0.28 0.06 0.13 0.04

Depression (R2 � 0.29)Support outside work �0.10 0.11 0.03 �0.01Supervisor support �0.09 0.07 0.03 �0.01Psychological job demands 0.03 0.12 0.01 0.01

Perceived stress (R2 � 0.33)Conflict at home 1.05 0.20 0.31 �0.01Imbalance 2.09 0.13 0.56 �0.01

Job dissatisfaction (R2 � 0.47)Supervisor support �0.19 0.26 0.03 �0.01Emotional load 0.20 0.12 0.05 �0.01

Role ambiguity 0.13 0.09 0.04 �0.01

* An alternative model for anger obtained when mental demands was restricted from entrybecause of its high correlation with other job stressors predicted 22% from three factors(semi-partial R2): low supervisor support (0.11), imbalance (0.06), and psychological jobdemands (0.05).

† An alternative but equally predictive model for fatigue was obtained when mentaldemands was restricted from entry yielded two predictors (semi-partial R2): low supervisorsupport (0.13) and psychological job demands (0.05).


job dissatisfaction levels were notexcessive, the results of this studysuggest that targeted efforts to facil-itate balance between work and per-sonal life, to reduce the prevalence ofselected job stressors and to increasesocial support may be important toenhancing the well-being and satis-faction of flight attendants.

These results provide the onlyknown empirical investigation ofchronic job stressors and their asso-ciation with psychological distressand job satisfaction levels amongflight attendants in the United States.Associations were in the hypothe-sized direction and were mainly con-sistent with those found in otherstudies of job stress among wom-en.11,15,16 One result that was unex-pected, however, was the positiverelationship between higher bodyweight and anxiety, since 99% of theparticipants were at or below levelsgenerally regarded to reflect ahealthy body weight (BMI � 25).Until the early to mid-1990s, mostUS airlines included body weightstandards among the criteria used tojudge flight attendants’ fitness forduty, and periodic weigh-ins wereused to demonstrate compliance(personal communication with Ms.Emily Carter, September 12, 2002,National Health Coordinator, Asso-ciation of Professional Flight Atten-dants). This historical industry prac-tice, combined with socialexpectations from the public, mayhave caused participants to be moreself-conscious and anxious abouttheir weight. It is also possible thatsocial desirability caused partici-pants to under-report their weight,although gross under-estimationseems unlikely.

Flight attendants with preschool-aged children were more likely toexperience difficulty establishingbalance between work demands andoutside obligations and “imbalance”was found to be associated with an-ger, anxiety, and perceived stress.These findings are consistent withdescriptive reports of job stress con-ditions of flight attendants in Eu-

rope6 and in the United States,3 in-cluding a targeted investigation ofhow flight attendants in the UnitedStates manage work and dependentcare responsibilities.17 Imbalance isa topic that has received increasingattention in the past decade as animportant issue for the health andwell-being of working women.18–21

Imbalance may be especially greatfor flight attendants who can face notonly irregular and long work hours,but also unpredictable schedulechanges and extended absences fromhome or loved ones. Expanding thesupport already provided by someairlines for extended care childcarefacilities near or at airports maygreatly aid flight attendants in bal-ancing their work and family rolesand reducing levels of distress.17,22

Emotional load levels were foundto be high, a result that was corrob-orated by qualitative results illumi-nating the sometimes critical natureof flight attendant responsibilities forpassenger safety (eg, medical emer-gencies in-flight), passenger assis-tance during flight delays, and occa-sional uncooperative and abusivepassenger behavior directed at flightattendants. In this study, emotionalload as well as emotional rewardswere investigated for their possibleassociation with distress, perceivedstress and level of job satisfaction.Neither variable was found to beassociated with psychological dis-tress or perceived stress, but emo-tional load was predictive of jobdissatisfaction. At the time of thisstudy, scales used to measure emo-tional labor and rewards were underdevelopment and had not yet beenfully validated.10 Reliability for thetwo-item emotional labor scale usedin this study was low (0.52), and asingle-item measure was used inplace of the scale to represent thisconstruct. Further research with animproved measure of emotional la-bor seems warranted to improve ourunderstanding of the potential role ofthis construct on distress levelsamong flight attendants.

Data for this cross-sectional studywere collected from a small sample(n � 73) of female flight attendantsemployed at two commercial airlinesin the United States, limiting thegeneralizability of our findings. Theprofile of job stressors in this occu-pational group as a whole could dif-fer from those reported here as aresult of varied employment policiesand benefit structures across the air-line industry. Large differences areunlikely, however, because opera-tional procedures between airlinesvary little due to regulatory mandatesand because all airlines would havefaced pressures to adopt similar em-ployment policies and benefit struc-tures to compete for flight crew em-ployees during the period of datacollection in 1995 and 1996 whenunemployment was at or near histor-ically low levels.

We should emphasize that data forthis study were collected before theevents of September 11, 2001 or“9/11” when the lives of all flightcrew and passengers on four hi-jacked commercial flights were trag-ically lost, along with more than2500 people who occupied targetedbuildings. Immediately followingthese incidents, efforts to interceptpossible additional hijackings led tothe suspension of all civilian aviationoperations for more than 2 days.Flight attendants working on 9/11were on board flights that wereaborted and diverted to alternate air-ports, leaving many in unfamiliarairports and cities for days and post-poning their reunion with concernedloved ones. Critical flying incidentshave been shown to contribute to thedevelopment of flying phobias andchronic psychiatric reactions amongflight crew members.1,2 Recent find-ings from a national survey on theeffects of 9/11 showed that psycho-logical distress reactions were beingexhibited not only among those withdirect exposure to the attacks butalso among the children living inthose households.23 It is important torecognize that all aircrew personneland their family members may be at


increased risk of adverse psycholog-ical reactions in the aftermath of 9/11.

The events of 9/11 underscore thereal potential for flight attendants tobe directly exposed to violence in theworkplace. Although terrorism wasspecifically mentioned by at leastone study participant as a factor thatmay place flight attendants healthand safety at risk, the quantitativemeasures used in this study focusedon chronic job stress conditions.Daily diary reports of unusual eventsduring flight were obtained amongmore than half of the participants,providing some information on thetype of acute stressor conditions thatmay occur more routinely in flightattendant work, such as uncoopera-tive or abusive passenger behavior,medical emergencies or accommoda-tions for passengers with specialneeds, and weather-related flight de-lays. Although diary reports minimizethe chance for recall bias, the openended diary design applied in thisstudy did not permit a systematic ex-amination of the frequency of specificunusual events. A future event diarysurvey, coupled with the monitoring ofselected biological indices of stress,would permit a more complete under-standing of the nature and pattern offlight attendants’ exposure to job stres-sors and aid in determining the acuteimpact of selected events for flightattendant health and well-being.

After examining the correlationsamong the conceptually distinct dis-tress measures, the individual dis-tress measures were retained to per-mit us to more fully characterize thepsychological distress profiles of theparticipants and to elucidate morespecifically the impact of job stres-sors on the flight attendants’ healthand well-being. While some inde-pendent variables were common pre-dictors of anger, anxiety, fatigue anddepression, some important distinc-tions emerged (eg, the associationsbetween BMI and anxiety, conflict athome and perceived stress, emo-tional load and job dissatisfaction).These distinctions are believed to berelevant to understanding the impact

of these predictors and, conse-quently, are relevant to prevention.

The moderately high levels of fa-tigue found in this study are consis-tent with other studies of flight atten-dants.3,24,25 This finding alsocorresponds with the objectivelymeasured sleep displacement andcircadian rhythm disruption mea-sures from study A (Grajewski et al.Measuring Circadian Rhythm Dis-ruption in Female Flight Attendants,submitted.).26 A recent study foundthat job stress characteristics such ashigh psychological job demands andlow supervisor support were associ-ated with fatigue among women.27

Moreover, misalignment of thesleep-wake cycle and circadianrhythms may be a chronic occupa-tional condition that may not be re-adjusted fully by rest after each dutyperiod. While associated fatigue lev-els may not impair job performance,feelings of fatigue and associatedbiological disruption may adverselyaffect the physical health of femaleflight attendants, including adversereproductive and cardiovascular out-comes.28–32 We are currently evalu-ating reproductive hormonal profilesof flight attendants in study A toinvestigate the association betweenreproductive health and circadiandisruption.

Depression levels were found tobe low among the flight attendants inthis study. This result was corrobo-rated by examination of data ob-tained on medication use in study A,which showed that only two of theforty-five (4%) flight attendants en-rolled reported current use of pre-scription anti-depressants (informa-tion about medication use amongparticipants of study B was not ob-tained). Self-selection factors mayresult in flight attendants havingmore favorable mental health thanthe general population, which mayexplain the low levels of depressionand the moderate to low levels ofother study outcomes. It is also pos-sible that higher outcome levels werenot found in this study because thesubstantial demands placed upon

participants in study A may haveresulted in more distressed and dis-satisfied flight attendants’ decliningparticipation.

Caution in the interpretation of themultiple regression findings is war-ranted due to the presence of expo-sure covariation and due to thestudy’s small sample size. It is im-portant to note that several of the jobstressor variables represented in thefinal models were correlated withother predictive job stressors. It maytherefore be the joint effect of thesestressors that is important for pre-dicting the study outcomes, morethan the effect of individual stres-sors. We therefore encourage readersto consider the collective results re-ported in this paper when interpret-ing the associations between jobstressors and psychological distress,perceived stress and job dissatisfac-tion.

The data from two feasibility stud-ies were combined for use in this jobstress investigation to increase statis-tical power. Data from study A wereobtained by self-administered ques-tionnaire and the data from study Bwere obtained by telephone interviewusing the same scale measures andresponse options. Although the modeof administration may have influencedthe response patterns of participants, acomparison of the two data sets re-vealed few differences in the meansand no differences in the direction ofthe relation (slope) between the inde-pendent and dependent variables.

Because statistical power was low,we were constrained in our ability toexamine differences among flight at-tendant groups. We were unable toexamine differences in job stressprofiles between flight attendants ac-cording to factors that have beenreported by others to be importantdeterminants of work load, such ascabin assignment (unpublished re-port), job seniority,3 and gender.5,33

A larger scale investigation is war-ranted to confirm differences in jobstressors between these factors toidentify flight attendants who may beat greater risk for distress.


The goal of this study was toevaluate the presence of chronic jobstressors and to examine the relation-ships between those stressors andpsychological distress, perceivedstress and level of job satisfaction.Standard questions and multi-itemscales were used to assess job stres-sors based upon flight attendants’responses. Since we sought to obtaininformed and impartial assessmentsof job conditions, devoid of flightattendants’ perceptions or appraisals,care was taken to select valid andreliable scale measures that were“low in their dependency on cogni-tive and emotional processing.”34

Although they are self-reported indi-cators themselves, flight attendants’daily diary reports of unusual eventsduring flight provide some basis forthe validity of the assessment of jobstress conditions in this study.

ConclusionsThe results of this study conducted

before the September 11, 2001 ter-rorist attacks suggest that certainchronic job stressors such as highmental and/or psychological job de-mands, imbalance between job de-mands and outside obligations, lowsupervisor support, and emotionalload have a substantive effect onpsychological distress, perceivedstress and job dissatisfaction amongflight attendants, following adjust-ment for individual factors. Despitemoderate to low levels of distressand dissatisfaction, targeted effortsto reduce modifiable job stressorsmay significantly improve the well-being and satisfaction of flight atten-dants. These results can be also beused to inform the content and de-sign of future studies of job stressamong flight attendants. These stud-ies should document the likelychanging levels of psychological dis-tress and job stress conditions toinform appropriate levels of em-ployee assistance.

AcknowledgmentsWe wish to thank the flight attendants who

participated in this study. We appreciate the

assistance of the study airlines and theirrespective unions, the Air Transport Associ-ation, and Ms. Jean Watson of the FederalAviation Administration. We wish to thankDr Ted Scharf of NIOSH and Dr CynthiaPilch for their consultation and assistance inselecting measures for this study. We alsoacknowledge the efforts of our field staff andtelephone interviewers: Barbara Jenkins,Chris Gersic, BJ Haussler, Pamela Anderson,Karen Davis-King, Connie Brouse, LyndaEwers, Relada Miller, Julie Toles, MarthaDiMuzio, Mimi Nguyen, and Laurie Pia-citelli.

Funding for this research was provided, inpart, by the Federal Aviation Administrationthrough an Interagency Agreement.

References1. Dyregrov A, Skogstad A, Helles�y OH,

Haugli L. Fear of flying in civil aviationpersonnel. Aviat Space Environ Med.1992;63:831–838.

2. Marks M, Yule W, De Silva P. Post-traumatic stress disorder in airplane cabincrew attendants. Aviat Space EnvironMed. 1995;65:264–268.

3. Smolensky MH, Lee E, Mott D, ColliganM. A health profile of American flightattendants. J Hum Ergol. 1982;11(Suppl):103–119.

4. Suvanto S, Ilmarinen J. Stress and strainin flight attendant work. Ergonomia.1989;12:85–91.

5. Skogstad A, Dyregrov A, Helles�y OH.Cockpit-cabin crew interaction: satisfac-tion with communication and informationexchange. Aviat Space Environ Med.1995;66:841–848.

6. Ballard T, Corradi L, Lauria L, MazzantiC, Sgorbissa F, Scaravelli G, Romito P.Flight attendants talk about work, familyand health: generating hypotheses fromqualitative research. In: Bildt C, Gonas L,Karlqvist L, Westberg H, eds. Women,Work and Health Book of Abstracts.Stockholm: National Institute for Work-ing Life; 2002:341.

7. Whelan EA, Grajewski B, Wood E, et al.Feasibility issues in reproductive bio-monitoring of female flight attendantsand teachers. J Occup Environ Med.2002;44:947–954.

8. Hurrell JJ, McLaney MA. Exposure tojob stress: a new psychometric instru-ment. Scand J Work Environ Health.1988;14:27–28.

9. Karasek R, Brisson, C, Kawakami N,Houtman I, Bongers P, Amick B. The JobContent Questionnaire (JCQ): an instru-ment for internationally comparative as-sessments of psychosocial job character-

istics, Occup Health Psychol. 1998;3:322–355.

10. Spratt K. Emotions at work. Dissertation:Johns Hopkins University, Baltimore,MD; 1994.

11. Piltch C. Work and mental distress: acomparative analysis of the experiencesof women and men. Dissertation: BostonUniversity, Boston, MA; 1992.

12. Cohen S, Kamarck T, Mermelstein R. Aglobal measure of perceived stress.J Health Soc Behav. 1983;24:385–396.

13. Rummel RJ. Applied Factor Analysis.Evanston: Northwestern UniversityPress; 1970:271–280.

14. World Health Organization TechnicalReport Series. Physical Status: the Useand Interpretation of Anthropometry. Ge-neva: World Health Organization; 1995.

15. Bildt C, Michelsen. Gender differences inthe effects from working conditions onmental health: a 4-year follow-up. IntArch Occup Environ Health. 2002;75:252–258.

16. Nelson DL, Burke RJ. Gender, WorkStress and Health. Washington DC:American Psychological Association;2002.

17. Desrosiers A, Emlen AC. Airlines, FlightAttendants, and Dependent Care. Port-land State University, Regional ResearchInstitute for Human Services, 1997.Available at: http://www.hhs.oregonstat-e.edu/familypolicy/occrp/publications/AFA-study.pdf.

18. Hochschild A. The Second Shift. NewYork: Avon Books; 1989.

19. Hall E. Double exposure: the combinedimpact of the home and work environ-ments on psychosomatic strain in Swed-ish women and men. Int J Health Serv.1992;22:239–260.

20. Barling J Work and family: in search ofmore effective workplace interventions.In Cooper and Rousseau, Eds. Trends inOrganizational Behavior, Volume 1.London: John Wiley & Sons; 1994.

21. Hewlett SA. Executive Women and theMyth of Having it All. Harv Bus Rev.2002;80:5–11.

22. Shellenbarger S. Flight attendants under-line care woes of overnight workers. WallStreet J. August 30, 1995.

23. Schlenger WE, Caddell JM, Ebert L, etal. Psychological reactions to terroristattacks. JAMA. 2002;288:581–588.

24. Bassett JR, Spillane R. Urinary cortisolexcretion and mood ratings in aircraftcabin crew during a tour of duty involv-ing disruption of circadian rhythm. Phar-macol Biochem Behav. 1987;27:413–420.

25. Haugli L, Skogstad A, Helles�y OH.Health, sleep, and mood perceptions re-


ported by airline crews flying short andlong hauls. Aviat Space Environ Med.1994;65:27–34.

26. Grajewski B, Whelan EA, Nguyen MM,Kwan LC, Cole RJ. Sleep disturbanceamong female flight attendants andteachers in a reproductive biomonitoringstudy. Am J Epidemiol. 2000;151:S82.

27. Bultmann U, Kant IJ, Schroer CAP, KaslSV. The relationship between psychoso-cial work characteristics and fatigue andpsychological distress. Int Arch OccupEnviron Health. 2002;75:259–266.

28. Okatani Y, Sagara Y. Enhanced noctur-nal melatonin secretion in women with

functional secondary amenorrhea: rela-tionship to opioid system and endoge-nous estrogen levels. Horm Res. 1995;43:194–199.

29. Brzezinski A. Melatonin in humans.N Engl J Med. 1997;336:186–195.

30. Theorell T. Psychosocial cardiovascularrisks—on the double loads of women.Psychother Psychosom. 1991;55:81–89.

31. Knutsson A, Hallquist J, Reuterwall C,Theorell T, Akerstedt T. Shiftwork andmyocardial infarction: a case-controlstudy. J Occup Environ Med. 1999;56:46–50.

32. Boggild H, Knutsson A. Shift work, riskfactors and cardiovascular disease. ScandJ Work Environ Health. 1999;25:85–99.

33. Hochschild A. The Managed Heart:Commercialization of Human Feeling.Los Angeles: University of CaliforniaPress; 1983.

34. Frese M, Zapf D. Methodological issuesin the study of work stress: objective vssubjective measurement of work stressand the question of longitudinal studies.In Cooper CL, Payne R, eds. Causes,Coping and Consequences of Stress atWork, Chapter 12. Chichester, UK:Wiley; 1988:375–411.

APPENDIX IJob Stressor Measures.

Job Stressors Questionnaire ItemsCronbach’sCoefficient*

MeasurementRange

Role ambiguity There are clear planned goals and objective for my job (r). [0.39] (1–7)a

Role conflict I have to bend or break a rule or policy in order to carry out anassignment.

[0.61] (1–7)a

Supervisor support Supervisor makes work life easier; is easy to talk to; can be re-lied upon.

0.85 (3–12)a

Co-worker support Co-workers make work life easier; are easy to talk to; can berelied upon.

0.71 (3–12)a

Mental demands Long periods of intense concentration required; tasks are ofteninterrupted requiring attention at a later time; waiting for othersslows me down.

0.64 (3–12)b

Psychological job demands My job requires working very fast. [0.49] (1–4)b

Decision latitude Sum of two subscales: 1) skill discretion (6-items: keep learning,requires creativity and skill, has variety, can develop abilities,repetitive work (r)); 2) decision authority (3-items: make deci-sions, a lot of say, little freedom to decide (r)).

0.70 (24–96)b

Job strain Computed from the ratio of the decision latitude and psycholog-ical job demand scales.

– (0.25–2.0)b

Task control Influence over amount of work; pace of work; work schedule;decisions about time sequence; ability of take a short break.

0.71 (5–25)a

Emotional load At work I have to seem concerned about passengers even whenI don’t feel like it.

[0.52] (1–5)c

Emotional rewards When customers say “thank you” it makes the job worthwhile;working with people is satisfying.

0.65 (2–10)c

Conflict at work I am exposed to hostility or conflict from the people I work with. – (1–4)e

Imbalance Conflicts exist between job demands and outside obligations;the demands of my job create serious stress in my family.

– (0–1)d

Job insecurity My job security is good (r); during the past year, did you everface job loss or layoff?

– (0–1)a

Source: a) NIOSH Generic Job Stress Questionnaire (Hurrell and McLaney, 1988), b) Job Content Questionnaire (Karasek et al., 1998), c)The Emotions at Work Scale, Version 1 (Spratt, 1994), d) composite measure adapted from Piltch (1992), e) measure adapted from theNorthwestern National Life Survey on Fear and Violence in the Workplace (1993). (r) � reverse coded

* Cronbach’s coefficients in brackets are for the multi-item scale version of the measure.


APPENDIX IIPsychological Distress, Perceived Stress, and Job Dissatisfaction Measures

Dependent Variables Questionnaire ItemsCronbach’sCoefficient

MeasurementRange

Anger Felt resentful; felt bad-tempered; felt angry. 0.79 (0–9)a

Anxiety Felt tense; felt nervous; felt on the edge. 0.79 (0–9)a

Fatigue Felt bushed; felt fatigued; felt worn-out. 0.89 (0–9)a

Depression Felt worthless; felt hopeless; felt sad. 0.68 (0–9)a

Perceived stress Felt things were going your way (r); felt confident in your ability to handle prob-lems (r); felt difficulties were piling up; felt you were unable to control the im-portant things.

0.74 (0–12)b

Job dissatisfaction All in all, how satisfied would you say you are with your job? (1 � very satisfied,4 � not at all satisfied)

– (1–4)c

Source: a) Profile of Mood States (McNair DM et al., 1981); b) Perceived Stress Scale (Cohen et al., 1983); c) NIOSH Generic Job StressQuestionnaire (Hurrell and McLaney, 1988).

(r) � reverse coded.


job stress among female flight attendants

Documents