/div>
otional expres-
si
r from the
pe
emotional stress or an environment laden
w
work burnout, the first
ergy and have the sense that they are becoming exhausted
time and to display appropria
pressions and behaviors. At the same time, the organi-
zations’ regulations and salary structure require that these
attendants control their emotions to create the work cli-
mate that the organization wants.” In order to meet the
organization’s expectations for client service, most em-
ployees must frequently control their own emotions in
such a way as to convey the emotion expected by the
organization.
Referring to
bor, Wharton [5] proposed three work characteristics or
perspectives:
1) The work
th face to face and conversational exchanges.
2) When working, the employee is expected to
rtain emotions that will have the desired influence on
others.
3) The
ited by employer-imposed regulations.
Goffman [6] argued that the form of em
on imposed by emotional labor amounted to surface
acting. Goffman suggested that emotional labor amounts
to a dramatic enactment, one similar to that of an actor
who is expected to express emotion in a screenplay.
Goffman [6] also posited that, in addition to surface act-
ing, workers may choose to modify how they think about
their own feelings or to resist conforming to the emo-
tional expression as dictated by the external organization.
Goffman’s model of emotional labor as drama su- ggests
that employees may also adopt “deep acting”.
Earlier researchers discussed emotional labo
rspectives of high- and low-level emotional labor as
proposed by Hochschild [4]. Discussion about these two
levels, however, does not completely address the emo-
tional labor load and its component factors. The four
perspectives proposed by Morris and Feldman [7] lacked
the support of empirical research. A test of Morris and
Feldman’s [8] approach discovered that, although their
model held promise for explaining the capacity for emo-
tional labor, it was unable to differentiate causality from
correlation in explaining emotional labor and its conse-
quences.
Long-term
ith too much emotional labor may cause emotional ex-
haustion, another consequence of emotional labor. When
employees have contact with clients for a long time, the
heavy emotional load may induce energy exhaustion [9].
This may cause employees to withdraw from their duties
and may lead to illness [10,11].
1.2 Emotional Exhaustion
According to the literature on
phase of work burnout is emotional exhaustion [12].
Emotional exhaustion occurs when individuals lack en-
Copyright © 2009 SciRes JSSM
Flight Attendants’ Emotional Labor and Exhaustion in the Taiwanese Airline Industry307
ustion as a result of
fre
ell as the organization. On the psychological
le
nd Procedure
first-line employees, i.e.,
omestic and international
H1: Flight attendants with different backgrounds have
[10]. Saxton, Phillips, and Blakeney [9] also pointed out
that emotional exhaustion is related to excessive emo-
tional demand during interpersonal interaction; this
causes energy exhaustion and increased withdrawal.
These conditions decrease productivity and increase cer-
tain stress reactions. In the same vein, Maslach and
Jackson [13] found that workers who had to engage in
face to face contact with clients as well as manage emo-
tional expression for extended periods were most suscep-
tible to emotional exhaustion.
Based on these findings, it seems probable that flight
attendants develop emotional exha
quent and extended interactions with customers. Those
who experience excessive interaction with the clients or
who obey the organization’s expectations for emotional
control [14] show increasingly frequent emotional ex-
haustion.
Emotional exhaustion harms an individual’s body and
mind as w
vel, emotional exhaustion leads to lower self-esteem,
depression, nervousness, and irritability. On the physio-
logical level, individuals might experience headaches,
insomnia, stomach upset, and so on [15].
2. Method
2.1 Sample a
The current research focused on
flight attendants, employed by d
airlines in Taiwan (Far Eastern Air Transport Corp.,
Mandarin Airlines, UNI Ai, TransAsia Airways, EVA
Air, and China Airlines). We distributed 500 question-
naires and received 380 responses. After eliminating de-
fective questionnaires, we were left with 353 question-
naires, for an effective response rate of 70.16%. Because
of time and resource limitations as well as the work cha-
racteristics of flight attendants and their variable sched-
ules, we delivered the questionnaires to selected senior
flight attendants and asked them to transmit the ques-
tionnaires to other attendants working for national and in-
ternational airlines through their own company and other
companies. We also asked these flight attendants to com-
plete the questionnaire when they were on duty. Because
the work time and airline schedules of flight attendants
are not fixed, and because the shift schedules are ar-
ranged by computer, flight attendants are subject to arbi-
trary posting by the airline that trained them. Therefore,
each flight attendant may work with a different flight
team at any given time and may work with the same
flight team over the short term only rarely. Thus, we pro-
vided the questionnaires without knowledge of which
flight attendants would receive them. Therefore, this
sampling method amounts to simple random sampling.
2.2 Hypotheses
significantly different feelings about emotional labor.
H2: Flight attendants with different backgrounds have
significantly different feelings about emotional exhaus-
tion.
H3: Flight attendants’ emotional labor is positively
correlated with emotional exhaustion.
H4: Flight attendants’ emotional labor is predictive of
emotional exhaustion.
2.3 Measures
2.3.1 Emotional Labor
The first part of the emotional labor
e “Organization Emotional Labor Scale”
specially for Taiwan, measures
lach
rvey (MBI-GS) developed
Scale development:
scale, cited as th
developed by Lin [16] e
emotional labor. It adopts a 5-point Likert scale to score
the items from 1 to 5 as absolute disagreement, disagr-
eement, neutral, agreement, and absolute agreement, re-
spectively. The higher the total score, the greater the em-
ployee’s level of emotional labor. The original emotional
labor scale had 24 questions divided into five themes:
basic emotional expression, surface acting, deep acting,
multiformity, and interpersonal interaction. Taking into
account the test of professional validation as well as sug-
gestions by experts, the descriptions of two questions in
the original scale were modified. Finally, the authors de-
signed the pre-questionnaire. Using SPSS, the authors
conducted an item analysis on responses to the
pre-questionnaire designed to test the scale questions.
This research will adopt, as indices of internal consis-
tency, the criterion method of discrimination analysis and
correlative analysis on questions and the total score on a
test of isomorphism type to analyze the pre-question-
naire.
Scale reliability and validation analysis: Analysis of
the internal consistency of question 16 showed that the
correlation coefficient of items and total scores was lower
than 0.3 and that the discrimination was unacceptably
low. Therefore, we deleted this question. The discrimina-
tion values of Question 15 and Question 17 were low,
and they did achieve the level of significance. Therefore,
the descriptions of these two questions were modified.
Based on the item analysis, the emotional labor scale
retained 23 questions. Cronbach’s α calculated for the
entire emotional labor scale was 0.883, and the reliability
the entire scale achieves was over 0.7. The reliability
value of the themes within the scale was calculated at 0.6,
which is within the acceptable range. These results dem-
onstrated the reliability of this emotional labor scale.
2.3.2 Emotional Exhaustion
Development of scale: The second part of the emotional
exhaustion scale cites the third edition of the Masch
Burnout Inventory-General Su
by Maslach and Leiter [17]. There are five questions in
this scale. This research adopts a 5-point Likert scale to
score the items from 1 to 5 as absolute disagreement,
Copyright © 2009 SciRes JSSM
Flight Attendants’ Emotional Labor and Exhaustion in the Taiwanese Airline Industry
Copyright © 2009 SciRes JSSM
308
addition, the critical
ra
. The demographic data were as follows: Age:
twelve flight attendants were 31-35 years
g for 31.7% of the sample. Marital status:
is
es included age, marital status,
ed
nalysis using Scheffe’s
te
Table 1. Abstraction of the analysis on domestic flight attendan
5-point Likert scale (N=353)
disagreement, neutrality, agreement, and absolute agree-
ment, respectively. The higher the total score, the higher
the degree of emotional exhaustion an employee feels.
The original scale was tested by expert validation and we
retained all six terms to develop the pre-questionnaire by
referring to the experts’ suggestions.
Scale reliability and validation analysis: Referring to
the item analysis results, the correlations of all questions
and the subscale of emotional exhaustion scale were
above 0.3, achieving significance. In
tios were higher than 3 and achieved significance.
Therefore, the five questions of the emotional exhaus-
tion scale are all retained, in keeping with results of the
item analysis. The Cronbach’s α for the entire emotional
exhaustion scale was 0.755, and the reliability of the
entire scale was more than 0.7. Both are in the acceptable
range, indicating that this emotional exhaustion scale is
reliable.
3. Results
The effective sample of flight attendants was 353 re-
spondents
One hundred
old, accountin
Two hundred attendants were single, accounting for
56.7% of the sample. Educational background: One hun-
dred seventy-one attendants had graduated from junior
college, accounting for 48.4% of the sample. Service year:
One hundred eighty-four attendants had served for three
to seven years, accounting for 51.1% of the sample. Av-
erage salary: One hundred sixty-two flight attendants
reported salaries ranging from NTD60,000 to NDT
70,000, accounting for 45.9% of the sample. The data
regarding flight attendants’ emotional labor and emo-
tional exhaustion are displayed in Table 1
There were 23 questions on the emotional labor scale.
Results showed an average emotional labor score of 4.15
(SD=0.392) on the 5-point Likert scale; the average score
thus moderately high. This indicates that the flight at-
tendants believe that they must obey the company’s
regulations regarding their emotional expression while
working. Results indicate that these flight attendants en-
gaged in a high level of emotional labor. Analyses of
each element of emotional labor indicated that interper-
sonal interaction scored highest on the emotional labor
scale (M=4.39, SD=0.569); next highest was basic emo-
tional expression (M=4.28, SD=0.482); multiformity
achieved the lowest score (M=3.66, SD=0.642). There
were five questions on the emotional exhaustion scale.
The average score over all questions was 3.81 (SD=.528).
This result indicates that the feeling of exhaustion ex-
perienced by flight attendants in response to emotional
labor is moderately high.
The current study examined whether different back-
grounds cause differences in feelings about emotional
labor. Background variabl
ucational background, service year, and average salary.
Results are displayed in Table 2:
Age: Flight attendants of different ages showed sig-
nificant differences in their perspectives on emotional
labor. According to the post hoc a
st, feelings regarding surface acting were experienced
more strongly by flight attendants 20-25 years of age, and
those older than 41, than by those 26-30.years old. Feel-
ings about deep acting were stronger among attendants
20-25 years old, 31-35 years old, and older than 41 years
old than among those 26-30 years old. Feelings of multi-
formity were more profound among flight attendants
31-35 years old than among 36-40 year-old attendants.
The degree of emotional exhaustion of flight attendants
36-40 years old was more pronounced than that among
those 20-30 years old. In addition, feelings of emotional
exhaustion were stronger among flight attendants 31-35
years old than among attendants 20-25 years old. And,
finally, flight attendants older than 41 years had the
strongest feelings about emotional exhaustion.
ts’ emotional labor and emotional exhaustion measured on a
Name Number of
Questions Minimal ValueMaximal ValueAverage of Question Standard Deviation
of Question
Entire Emotional Labor 2.96 5.00 23 4.15 0.392
Basic Emotional Expression 2.86 5.00 7 4.28 0.482
Surface Acting 3 2.33 5.00 4.13 0.598
Deep Acting 7 2.71 5.00 4.15 0.531
Multiformity 3 1.67 5.00 3.66 0.642
Interaction 3 3.00 5.00 4.39 0.569
Entire Emotional Exhaustion 5 2.80 5.00 3.81 0.528
Flight Attendants’ Emotional Labor and Exhaustion in the Taiwanese Airline Industry309
Analysis of the difference beeen the basic attribute of rh object and research variable (n=353) Table 2. tw esearc
Ag
Item e Marital StatusEducational Background
(F Value) (F Value) (F Value) Service Year Average Salary
(F Value) (F Value)
Entire Emotional Labor 4. 004*2.445 0.531 2.248 2.928*
Basic Emotional Expression 0.771 1.621 0.527 1.729 3.776**
Surface Acting 4.868** 0.462 1.011 2.780* 4.745*
Deep Acting 9.083*** 4.928** 0.204 3.071* 2.636*
Multiformity 3.119* 0.290 2.620 1.552 0.798
Interaction 2.120 3.180* 0.454 3.441* 1.940
Entire Emotio1 6.290***7 .798**6. 355***10* .803**13* .340**
nal Exhaustion
*ificant level; **p < 0.01 extra significant level; *** p < 0.001 absolutely significant lp < 0.05 signordinaryevel
Tab on tion emotional the spece aspotional
exh le 3. Abstract of the analysishe correlatbetweenlabor andtrum of thects of em
austion (N=353)
1. 2. 3. 4. 5. 6. 7.
1. Spectrum of Emotional La b o r
2. Basic Emotional Expression 0.839 **
.501 ** .528 **
.110* .114 *
.412 ** .222 *
3. Surface Acting 0.751 ** 0.625 **
4. Deep Acting 0.765 ** 00
5. Multiformity 0.338 ** 0.032 00
6. Interaction 0.729 ** 0.624 ** 0.416 ** 00
7.Entire Level of Emotional Exhaust i on 0.256 * 0.138** 0.138** 0.371 * -0.055 0.211**
Data source: Integrated by this research; * P<0.05; ** Ple
was m
ants. Additionally, the emotional exhaustion of flight
at
than among those who had graduated
fr
d fewer
th
the overall level of emotional labor than did
fli
expressed by those with an average salary of
NTD80-90,000. Feelings of engaging in deep acting were
stronger among flight attendants with an average salary
kground variables, are both par-
tia
the elements of emotional labor, only
de
<0.01 (Sing tail)
Marriage: The deep acting of single flight attendants
ore pronounced than that of married flight atten-
d
tendants who selected the term “other” to describe their
marital status was stronger than that reported by married
flight attendants.
Educational background: The feeling of emotional
exhaustion among flight attendants who had masters de-
grees was greater
om high school, junior college, or college.
Service years: Flight attendants who had served for
three to seven years reported stronger feelings of emo-
tional exhaustion than did those who had serve
an three years. Correlatively, the feeling of emotional
exhaustion reported by flight attendants who had served
more than 15 years was more pronounced than that re-
ported by flight attendants who had served for three to
11years.
Average salary: Flight attendants with an average sal-
ary of NTD90-100,000 expressed more marked feelings
regarding
ght attendants with an average salary of NTD60-90,000.
Feelings regarding the basic emotional expression ex-
pected of flight attendants with an average salary of
NTD90-100,000, in turn, were more marked than those
of NTD90-100 than among those whose average salary
was less than NTD60,000. Finally, feelings about emo-
tional exhaustion expressed by flight attendants whose
average salary was NTD90-100,000 were stronger than
those reported by flight attendants with an average salary
of less than NTD60,000, and those with an average salary
of NTD60-90,000.
In combination, these results indicate that flight atten-
dants from different backgrounds have significantly dif-
ferent experiences regarding emotional labor and emo-
tional exhaustion. Thus, Hypothesis 1 and Hypothesis 2,
which posited that existence of such differences in ex-
perience based on bac
lly supported.
Referring to Table 3, the measure of overall emotional
labor showed a significant positive correlation with each
element (P<0.01) within the emotional labor measure,
and the score for overall emotional labor showed a sig-
nificant positive correlation with emotional exhaustion
(p<0.05). Among
ep acting showed a significant positive correlation with
emotional exhaustion (p<0.05). The other elements did
Copyright © 2009 SciRes JSSM
Flight Attendants’ Emotional Labor and Exhaustion in the Taiwanese Airline Industry
310
Table 4. Abstract of the stepwise multiple regression analysison each aspect of emotional labor for emotional exhaustion
(N=353)
Selective Sequence of R2 after
Variables
Multiple Correlation
Coefficient (R)
Determination Original Regression Standardized Regression
F Value
Coefficient (R)
2Regulation Coefficient (β) Coefficient
Intercept (Constant Term) 2.526
Deep Acting 0.371 0.137 0.135 5
Multiformity 0.383 0.147 0.142 30.141***-0.081 -0.099*
5.871***0.380 0.382***
*p < 0.05 significant level; *** p < 0.001 absolutely significant level
not reance. The correlations amthese
ificant positive correlation with each element of emo-
relation with over-
al
abor) predicted criterion variables in the re-
gr
r can be described
rt scale, with the average score being
ttendants’ average feeling about the
t ahad face contact with clients, and
had conversations for longer periods, and with higher
frequency. Therefore, flight attendants are also first-line
service people in the airline industry [1]. Adleman’s re-
fe
s have deeper feelings
ach significong
perspectives are described as follows:
The score for overall emotional labor showed a sig-
n
tional labor (P<0.01). Furthermore, overall emotional
labor showed a significant positive cor
l emotional exhaustion (p<0.01). Among the elements
of emotional labor, only multiformity failed to show a
significant correlation with emotional exhaustion ( r =
-0.055, p=0.15>0.05); the other elements of emotional
labor all showed significant correlations with emotional
exhaustion. Finally, the overall measure of emotional
labor showed a significant positive correlation with the
overall measure of emotional exhaustion. Thus, Hypothe-
sis 3, which predicted just such a relationship, was also
supported.
Referring to the stepwise multiple regression analysis
shown in Table 4, the two perspectives labeled deep act-
ing and multiformity (from among the five elements of
emotional l
ession model. The multiple correlation coefficient, R,
was 0.383, and its united explanation variance was 0.147,
indicating that these two variables explain 14.7% of the
entire emotional exhaustion measure. Referring to the
results above, deep acting and multiformity have signifi-
cant positive and negative predictive capability for emo-
tional exhaustion, respectively. Therefore, Hypothesis 4,
which posited that flight attendants’ emotional labor
would have a significant predictive effect on emotional
exhaustion, was also partly supported.
4. Conclusions
Referring to Table 1, it appears that domestic flight at-
tendants’ feelings about emotional labo
using a 5-point Like
4.15. Thus, flight a
role of emotional labor in their lives is moderately strong.
This result indicates that flight attendants can be classi-
fied as workers who engage in a high level of emotional
labor. Regarding feelings about various forms of emo-
tional labor, the element of emotional labor having to do
with interaction showed the strongest effect in the present
findings. Compared to colleagues in other departments,
search [18] mentioned that worker performance and sat-
isfaction were lower among those engaging in high levels
of emotional labor than among workers with less emo-
tional labor. High levels of emotional labor can easily
generate feelings of unhappiness, lack of self-respect, and
depression. Therefore, aviation management organiza-
tions must consider whether long-term emotional labor
will have negative emotional effects on flight attendants.
Flight attendants’ feelings about emotional exhaustion
can also be described using a 5-point Likert scale, with an
average score of 3.81. This indicates that, on average,
flight attendants in this study experienced a moderate
level of emotional exhaustion. More specifically, exami-
nation of the detailed questions regarding emotional ex-
haustion revealed that the highest average score was
given to the statement, “The whole work day makes me
flighttendants ce to fa
el tired.” this was followed in importance by the state-
ment, “My work makes me feel tired emotionally.” Thus,
most flight attendants reported that they felt tired when
working for a long period of time. This research high-
lights the consequences of the tiring nature of flight at-
tendants’ working environment. Take offs and landings
are frequent for domestic airlines, which means that flight
attendants must repeatedly provide customer service.
Thus, it is easy for flight attendants to feel that their work
is dull and repetitive. Flight attendants serving interna-
tional airline routes have to face a variety of passengers
with distinctive needs; thus, the company is stricter about
the entire range of services. All of these factors contrib-
ute to flight attendants’ feelings of exhaustion when
working for a long period of time.
Referring to Table 2, findings addressing differences
among flight attendants with different background vari-
ables and the relationships of these variables to emotional
labor and emotional exhaustion indicate that all back-
ground variables play a role. Relationships between emo-
tional labor and variables of age, marital status, service
years, and average salary all reached significance. Flight
attendants with longer service year
Copyright © 2009 SciRes JSSM
Flight Attendants’ Emotional Labor and Exhaustion in the Taiwanese Airline Industry311
ab
ut and workplace
ommercializa-
of service
The Presentation of self in everyday life,
abor,” Academy
axton, J. S. Phillips, and R. N. Blakeney, “The
. Dougherty, “A review and inte-
Schwab, and R. S. Schuler, “Toward an
ood
Organizational Behavior,
ences the course dis-
s thesis, 1999.
burnout,”
f
out emotional labor. This may be because service year
is directly related to a particular level of customer service.
New attendants may mainly serve in economy class,
those with intermediate service years may serve business
class, and those with the most service years may work in
first class. Because the passengers in first class have high
social status, the flight attendants must be especially at-
tentive; they have to concentrate on their movements,
emotional expressions, and so on. Therefore, we suggest
that this is the reason that flight attendants with longer
service years have deeper levels of emotional labor.
For feelings of emotional exhaustion among flight at-
tendants with different population backgrounds, relation-
ships between emotional exhaustion and the population
variables of age, marital status, educational background,
service years, and average salary all achieved signifi-
cance. As regards educational background, our findings
indicate that the higher the educational background of the
flight attendant, the higher are the requirements for their
w
[2]
ork performance. Cordes and Dougherty [10] pointed
out that high work achievement motivation and serious
attention to work can result in high stress and can easily
generate emotional exhaustion.
For the relationship between domestic flight atten-
dants’ emotional labor and emotional exhaustion, overall
emotional labor scores showed a significant positive cor-
relation with emotional exhaustion. This means that the
stronger the feelings about emotional labor, the higher the
emotional exhaustion. As regards the predictive effect on
emotional exhaustion of the various elements of emo-
tional labor, deep acting had the strongest value for pre-
dicting emotional exhaustion. This means that the
stronger the feeling about emotional labor, the higher the
level of emotional exhaustion. In particular, when flight
attendants realize that the company always responds to
customer complaints by supporting the client rather than
the employee, the degree of deep acting is likely to in-
crease. Therefore, airlines should consider carefully how
to decrease their employees’ feeling that they must en-
gage in deep acting; they should also consider more
carefully their treatment of employees, and whether they
should be more understanding and tolerant of employees.
REFERENCES
[1] S. J. Yuan, “The steward obeys the positive research of
hardworking ability and passenger’s satisfaction depend-
ence,” Department of Marketing and Distribution Man-
agement National Kaohsing First University of Science
and Technology Master's thesis, 2005.
H. C. Beh and P. McLaughlin, “Effect of long flights on
the cognitive performance of air crew,” Perceptual and
Motor Skills, Vol. 84, pp. 319–322, 1997.
[3] S. C. Liang and A. T. Hsieh, “Burno
deviance among flight attendants in Taiwan,” Psycho-
logical Reports, Vol. 101, pp. 457–468, 2007.
[4] A. R. Hochschild, “The managed heart: C
tion of human feeling,” Berkley, University of California
Press, 1983.
[5] A. S. Wharton, “The affective consequences
work: Managing emotions on the job,” Work and Occupa-
tions, Vol. 20, No. 2, pp. 205–232, 1993.
[6] E. Goffman, “
Doubleday and Co,” New York, NY, 1959.
[7] J. A. Morris and D. C. Feldman, “The dimensions, ante-
cedents, and consequences of emotional l
of Management Review, Vol. 21, No. 4, pp. 986–1010,
1996.
[8] M. J. Morris and D. C. Feldman, “Managing emotions in
the workplace,” Journal of Managerial Issues, Vol. 9, No.
1, pp. 257–274, 1997.
[9] M. J. S
changing nature of work,” San Francisco: Jossey-Bass, pp.
365–410, 1991.
[10] C. L. Cordes and T. W
gration of research on job burnout,” Academy of Man-
agement Review, Vol. 18, No. 4, pp. 621–656, 1993.
[11] S. E. Jaskson, R.
understanding of the burnout phenomenon,” Journal of
Applied Psychology, Vol. 71, No. 1, pp. 630–640, 1986.
[12] C. Maslach, “Burnout: The cost of caring,” Englew
Cliffs, NJ: Prentice-Hall, 1982.
[13] C. Maslach and S. E. Jackson, “Maslach burnout,” San
Francisco, CA: Jossey-Bass, 1986.
[14] A. Rafaeli and R. I. Sutton, “The expression of emotion in
organizational life,” Research in
Vol. 11, No. 1, pp. 1–42, 1989.
[15] L. C. Wang, “Role’s pressure influ
cussion that mood exhausts,” Department of Business
Administration, Institute of Behavior Medicine, National
Cheng Kung University, Master’
[16] S. P. Lin, “Organize the development of the form of
amount of load of labor service of mood,” Sun Yat-Sen
Management Review, Vol. 8, No. 3, pp. 427–447, 2000.
[17] C. Maslach and M. P. Leiter, “The truth about
San Francisco, CA: Jossey-Bass, 1997.
[18] P. K. Adelmann, “Emotional labor and employee
well-being,” Doctoral Dissertation, The University o
Michigan, 1989.
Copyright © 2009 SciRes JSSM