Академический Документы
Профессиональный Документы
Культура Документы
JOURNALBack
OF/ HOSPITALITY
COMMUNICATION SATISFACTION
& TOURISM RESEARCH
A FACTOR-ANALYTIC STUDY OF
COMMUNICATION SATISFACTION
IN THE LODGING INDUSTRY
Daniel J. Mount
Ki-Joon Back
The Pennsylvania State University
Communication, job satisfaction, and customer satisfaction have been popular topics in
hospitality and business research. The most widely used business communication satisfac-
tion questionnaire, the Communication Satisfaction Questionnaire (CSQ), has not been
used in the hospitality industry. This article presents a factor-analytic study of the CSQ in a
lodging setting. After the factor analysis, the article will then discuss the findings of com-
munication satisfaction as they compare to findings in other business-related studies.
Journal of Hospitality & Tourism Research, Vol. 23, No. 4, November 1999, 401-418
© 1999 Council on Hotel, Restaurant and Institutional Education
401
of the CSQ and its popularity, Downs and Hazen (1977) recommend that
researchers factor-analyze the scale to confirm the dimension structure.
Communication is critical in the hospitality industry. Understanding commu-
nication satisfaction, with its link to job satisfaction, should provide an ability to
better target resources to improve communication satisfaction issues. The CSQ is
the most widely used communication satisfaction questionnaire in business, and
yet no research has been noted that used the CSQ in hospitality. The CSQ should
provide a framework in which to study, measure, and address communication sat-
isfaction issues in the hospitality industry.
The primary objective of this article is to validate the CSQ as proposed by
Downs and Hazen to determine whether the CSQ can be used in research concern-
ing hotels. A review of the relationship of communication and job satisfaction will
be provided first, followed by a discussion of the theoretical structure of the CSQ.
A factor analysis will be completed with a discussion of the results. Once the CSQ
has been validated or modified as necessary, the results of this study can be com-
pared to results of studies in other fields. This research may then provide a guide
for further examination of the relationship between communication satisfaction
and job satisfaction in hospitality industries.
BACKGROUND
Communication
Organizational communication has been defined in research in a number of
different ways (Pincus, 1986). Goldhaber (1983) identified two major research
perspectives, process and perception. Information flow is the main concern for the
process perspective, whereas attitude or perception is the main concern for the
perception perspective. Both areas offered valuable contributions for understand-
ing structural properties of organizational communication systems (Pincus,
1986).
From a process perspective, Andrews and Herschel (1996) stated that informa-
tion could flow in three directions within organizational communication systems,
namely: downward, upward, and horizontally:
flows should be analyzed within the organization. In addition, it is clear that the
communicating organization must communicate with all of its employees at every
level whether the flow is downward or upward. Furthermore, Burns and Stalker
(1966) argued that organizations should have a flexible, or organic structure that
allows communication across various organizational departments and hierarchi-
cal levels such as open structures or upward systems.
The second perspective in the organizational communication research is the
perception perspective. Downs (1988) developed and defined the concept of com-
munication satisfaction as one emerging construct emphasizing the perception
perspective. Communication satisfaction is simply defined as satisfaction with
communication that is linked with the employee’s position in the organization.
Downs and Hazen (1977) stated that “communication satisfaction is multidimen-
sional rather than unidimensional” and set out to determine how the individual
dimensions relate to global job satisfaction. Therefore, this multidimensional
construct has been defined as “a summary of an individual’s satisfaction with
information flow and relationship variables” (Pincus, 1986).
role ambiguity and role conflict. In both cases, employees are unsure or confused
about the requirements of their job, which leads to job dissatisfaction.
For employees, needs satisfaction refers to the satisfaction of both economic
(i.e., wages) and noneconomic needs (i.e., internal service quality). Hallowell et al.
(1994) suggested that satisfaction with internal service quality is more powerful
than satisfaction with wages and benefits in predicting job satisfaction. They
regressed several dimensions of internal service quality on job satisfaction and
found significant relationships with goal alignment, management, teamwork, and
communication. By focusing on these and other components of internal service
quality, managers may develop their employees’ satisfaction toward their jobs,
which in turn may result in lower unintentional turnover and improved internal
and external service quality.
6. Corporate Information: deals with the broadest kind of information about the or-
ganization as a whole. It includes items on notification about changes, information
about the organization’s financial standing, and information about the overall poli-
cies and goals of the organization.
7. Personal Feedback: concerns what employees need to know about how they are
judged and how their performance is appraised. This deals with feedback in some
formalized setting or information that should be expected to be passed from supervi-
sor to subordinate.
8. Subordinate Communication: focuses both on upward and downward communica-
tion with subordinates. Only workers in a supervisory capacity respond to these
items, which include subordinate responsiveness to downward communication and
the extent to which subordinates initiate upward communication.
The coefficient alpha for the 510-employee test was .94. Coefficient alphas for
the eight dimensions have been consistently high, ranging from .72 to .96 for stud-
ies in the United States (Potvin, 1991/1992). The eight-factor structure has been
confirmed in numerous studies (Clampitt & Girard, 1987; Crino & White, 1981;
Pincus, 1986). Evidence of concurrent validity exists in that CSQ factors have
been found to be highly correlated with job satisfaction (Downs & Hazen, 1977),
to predict organizational commitment, and to be related to turnover (Clampitt &
Downs, 1993).
Clampitt and Downs (1993) feel that the most theoretical contribution of the
CSQ is the suggestion that communication satisfaction is a multidimensional con-
struct as opposed to a unidimensional one. It appears that, in the studies cited
above, the factors of Supervisory Communication and Subordinate Communica-
tion are the areas of the greatest employee satisfaction, whereas Personal Feed-
back provides the least satisfaction. In general, the research has suggested a rela-
tionship between job satisfaction and communication satisfaction (Clampitt &
Girard, 1993; Lee, 1989; Varona, 1996). Both Downs and Hazen (1977) and
Downs, Clampitt, and Pfeiffer (1988) found that three of the factors—Personal
Feedback, Communication Climate, and Supervisory Communication—have
been most strongly correlated with job satisfaction measures.
Greenbaum et al. (1988) suggested several advantages to using the CSQ. First,
the CSQ provides a relatively short and understandable instrument that can be
completed within 15 minutes. Second, scoring can be done easily using the Statis-
tical Package for Social Science (SPSS). Third, the instrument can be easily modi-
fied for various types of organizations. Last, the CSQ can measure more than one
outcome variable, such as job satisfaction and productivity. Although there are
several positive attributes associated with the CSQ, there is also one limitation:
Items dealing with interdepartmental communication and top management com-
munication are not within the dimensional structure (Greenbaum et al., 1988).
METHOD
The process of determining whether a model and the constructs that compose it
are valid is accomplished by investigating construct validity. To the extent that a
variable is abstract rather than concrete, it is spoken of as being a construct. Such a
Data Collection
The main components of the CSQ were used as presented by Downs and
Hazen (1977). Wording was changed on questions to make the questions more
understandable to hotel employees. Some of the demographic information was
modified or changed to fit the research situation. Questions 4 to 39 (36 ques-
tions) on the questionnaire composed seven of the eight proposed factors. Ques-
tions 42 to 46, answered only by management, composed the eighth factor, sub-
ordinate communication.
The questionnaires were administered at six hotels managed by the same com-
pany, located in South Texas. There was a large Hispanic employee population so
the questionnaires were translated into Spanish. The translation was completed by
a double-blind process. One person translated the original English CSQ into
Spanish. A second person took the Spanish-translated version and translated that
back to English. The translated English version was compared to the original to
determine if the original content of the question was retained. The questionnaire
used in this study is presented in the appendix.
RESULTS
The first step in the data analysis was to perform a reliability analysis. The
alpha for the 40-item instrument was .96, similar to the .94 alpha reported by
Downs and Hazen (1977). It was determined that the deletion of one question,
No. 28, would raise the alpha to .97. This question, “The grapevine is active in our
organization,” was found to be confusing even to the English respondents. There
is a positive connotation in all questionnaire items but, as phrased on this ques-
tionnaire, this question did not provide the positive connotation anchor. It was also
determined that the slang grapevine did not translate well into Spanish. This ques-
tion was removed from the analysis at this point.
A variety of rules have been suggested for determining the sample size
required to produce a stable solution when performing factor analysis. Many
researchers typically recommend that the necessary sample size be determined as
a function of the observed variables (Gorsuch, 1983; Nunnally, 1978). Guadag-
noli and Velicer (1988) studied the various rules in a Monte Carlo procedure. They
concluded that component saturation (the magnitude of factor loadings) and abso-
lute sample size were of major importance. They stated that the N-to-observed
variable rules were clearly not substantiated. The proposed factor analysis in this
study had 40 variables and eight factors, an average factor loading of .64. Based
on their suggestions, the minimum sample size is 150. Thus, the actual sample
size of 374 exceeds the suggested minimum.
The next step was to perform a confirmatory factor analysis based on the pro-
posed eight-factor structure of the original CSQ. The results of the confirmatory
analysis are presented in Table 2. To aid in the visual analysis, questions were
grouped by proposed factors. The actual question number from the survey is pre-
ceded by a two-letter designation of the proposed factor. The designations are as
follows: OI (organizational integration), PF (personal feedback), CI (corporate
information), CC (communication climate), SC (supervisor communication),
MQ (media quality), CW (coworker communication), and SB (subordinate com-
munication). Only the greatest factor loading for each variable is shown. Hair,
Anderson, Tatham, and Black (1998) suggest that factor loadings less than .30 are
not considered practically significant, and that factor loadings of .30 are statisti-
Table 1
Response Information
cally significant if the sample size exceeds 350. The sample size for this study was
374, so factor loadings of .30 or greater are shown in Table 2.
As can be seen in Table 2, the eight-factor structure did not confirm as pro-
posed. Although some factors, such as Communication Climate and Subordinate
Communication, held together well, in many, questions were separated over three
factors. The next step was to perform an exploratory factor analysis. As was dis-
cussed, a principal components extraction method analysis was completed first to
identify the number of factors in the factor solution. The number of factors was
identified through the use of a scree plot. Then, the number of factors was input
back into a common factor extraction method analysis. The scree plot of the prin-
cipal components extraction is shown in Figure 1.
As can be seen, the number of factors falling above the straight line was seven.
The selection of the seven factors for use in the common factor analysis is sup-
ported by other selection criteria. Only seven factors had eigenvalues greater than
1, and the cumulative percentage of variance explained by the seventh factor
exceeded 75%. The results of the common factor analysis are presented in Table 3.
Again, only loadings greater than .30 are presented.
Two of the proposed factors, Subordinate Communication and Coworker Com-
munication (without question No. 28), remain as originally constructed. Factor 1
had three variables from the Supervisor Communication factor and two variables
from the Media Quality factor. The proposed Supervisor Communication factor
seemed to separate based on the direction of the communication. The three vari-
ables that remained in Factor 1 all seemed to deal with a downward aspect of the
communication. The two Supervisor Communication variables that did not load
into Factor 1 seemed to deal more the supervisor listening to the employee, repre-
senting an upward or vertical flow of information. It is interesting to note that the
original definition found on page 6 defined Supervisory Communication in terms
of both upward and downward communication. The respondents in this study
seemed to differentiate between the two directions of communication. The two
Media Quality variables that loaded on Factor 1 addressed downward communi-
cation aspects that are controlled by the supervisor, clarity of written directives and
organization of meetings. This factor was renamed Downward Communication.
Table 2
Confirmatory Factor Analysis of the Communication Satisfaction Questionnaire
Factor
1 2 3 4 5 6 7 8
OI 4 .73
OI 5 .40
OI 10 .39
OI 11 .73
OI 15 .54
PF 7 .49
PF 8 .74
PF 9 .76
PF 14 .42
PF 18 .49
CI 6 .42
CI 12 .79
CI 13 .63
CI 16 .66
CI 17 .56
CC 19 .50
CC 21 .57
CC 23 .54
CC 26 .44
CC 27 .63
SC 20 .78
SC 22 .60
SC 25 .79
SC 29 1.00
SC 34 .72
MQ 24 .44
MQ 33 .69
MQ 35 .89
MQ 36 .72
MQ 38 .45
CW 30 .88
CW 31 .76
CW 32 .35
CW 37 .57
SB 42 .90
SB 43 .95
SB 44 .75
SB 45 .73
SB 46 .63
Note: The actual question number from the survey is preceded by a two-letter designation of
the proposed factor. The designations are as follows: OI (organizational integration), PF (per-
sonal feedback), CI (corporate information), CC (communication climate), SC (supervisor
communication), MQ (media quality), CW (coworker communication), and SB (subordinate
communication). Only the greatest factor loading for each variable is shown.
Factor 2 had seven variables. Four of the variables, representing the three larg-
est factor-loading scores, were from the original Corporate Information factor. It
is best to look at the three other variables in a Corporate Information perspective.
Figure 1
Scree Plot for Exploratory Factor Analysis
Question No. 24, from the original Media Quality factor, dealt with the quality of
the company’s publications. Question No. 5, from the original Organizational
Integration factor, dealt with the reception of information about personnel. Ques-
tion No. 14, from the original Personal Feedback factor, dealt with receiving
information about how problems on the job were being handled. It is possible to
understand how these questions could be interpreted in an Organizational Infor-
mation perspective. The factor name, Corporate Information, was retained for this
factor.
Factor 4 had eight variables. Three of the variables were from the original
Communication Climate factor, two variables were from the Organizational Inte-
gration factor, and one variable each was from the Personal Feedback, Supervi-
sory Communication, and Media Quality factors. A review of the variables not
from the original Communication Climate factor indicates that the factor may still
be interpreted from a Communication Climate perspective. For example, question
No. 24, from the original Media Quality factor, dealt with the appropriate amount
of communication in the company. The factor name, Communication Climate,
was retained for this factor.
Factor 5 had six variables, three from the original Personal Feedback factor.
Two of the other variables, questions No. 6 and 10, dealt with receiving informa-
tion about both departmental and corporate policies and goals. Question No. 4,
from the original Organizational Integration factor, dealt with receiving informa-
tion about progress in their job. It was surprising to see that this was not in the
Table 3
Exploratory Factor Analysis of the Communication Satisfaction Questionnaire
Factor
1 2 3 4 5 6 7
OI 4 .67
OI 5 .43
OI 10 .31
OI 11 .86
OI 15 .62
PF 7 .46
PF 8 .69
PF 9 .70
PF 14 .58
PF 18 .62
CI 6 .36
CI 12 .81
CI 13 .74
CI 16 .57
CI 17 .72
CC 19 .67
CC 21 .36
CC 23 .37
CC 26 .50
CC 27 .43
SC 20 .61
SC 22 .49
SC 25 .86
SC 29 .87
SC 34 .77
MQ 24 .35
MQ 33 .76
MQ 35 .87
MQ 36 .48
MQ 38 .59
CW 30 .91
CW 31 .87
CW 32 .30
CW 37 .65
SB 42 .81
SB 43 .90
SB 44 .68
SB 45 .82
SB 46 .67
Note: The actual question number from the survey is preceded by a two-letter designation of
the proposed factor. The designations are as follows: OI (organizational integration), PF (per-
sonal feedback), CI (corporate information), CC (communication climate), SC (supervisor
communication), MQ (media quality), CW (coworker communication), and SB (subordinate
communication). Only the greatest factor loading for each variable is shown.
Table 4
Factor List and Model Comparison Summary Information
Factor 7 had four variables. Although the variables do seem to capture several
concepts, there was one variable that loaded very strongly on this factor. That vari-
able was question No. 29, from the original Supervisory Communication factor,
which dealt with the supervisor being open to ideas. This factor was named Verti-
cal Communication. Table 4 presents a list of the original factors and the factors
identified in this study along with the summary model comparison information. It
can be seen that the summary information generally supports the use of the new
factor structure. The factor intercorrelation average was slightly higher for the
new factor structure; this may be because of the reduced number of factors.
The new factor structure was then used to compare the results of communica-
tion satisfaction relationships to job satisfaction between this study and other pre-
viously cited studies. In a study of the CSQ, Clampitt (1993) suggested that it
appeared that the factors of Supervisory Communication and Subordinate Com-
munication were the areas of the greatest employee satisfaction whereas Personal
Feedback provided the least satisfaction. The means of the new factors, in
descending order, in this study were: Downward Communication (7.39), Subordi-
nate Communication (7.39), Coworker Communication (7.00), Vertical Commu-
nication (6.86), Communication Climate (6.56), Personal Feedback (6.21), and
Corporate Information (5.85). In that the largest number of variables of the new
factor, Downward Communication, were from the Supervisory Communication
factor, the results support the suggestions of Clampitt (1993). In this study, Per-
sonal Feedback was the second lowest in terms of satisfaction.
Downs and Hazen (1977) and Downs et al. (1988) found that three of the fac-
tors—Personal Feedback, Communication Climate, and Supervisory Communi-
cation—have been most strongly correlated with job satisfaction measures. The
correlation coefficients of the new factors with overall job satisfaction, in
descending order, were: Downward Communication (.575), Communication Cli-
mate (.560), Vertical Communication (.535), Personal Feedback (.502), Corpo-
rate Information (.468), Coworker Communication (.415), and Subordinate Com-
munication (.319). All of the correlation coefficients were significant at .01 with
the exception of Subordinate Communication, which was significant at .05.
Again, remembering that the new Downward Communication is very much asso-
ciated with the old Supervisory Communication, these results again support the
findings of Downs and his colleagues.
CONCLUSION
As Downs and Hazen (1977) recommend, the use of the CSQ, or any other fac-
tor study, should be preceded by a factor analysis. The factor analysis presented in
this research indicates changes to the factor structure of the proposed CSQ.
Although the factor structure changed, the underlying theoretical structure of the
CSQ was supported. Factors that were the same as, or similar to, those factors
found least satisfying in previous research were identified in this research. Also,
factors that were found to be most correlated to job satisfaction were also similar
to factors proposed in previous research.
Future research should continue to explore the factor structure of the CSQ in
hospitality settings. Also, it would appear that more focused research can be con-
ducted on specific factors. From the factors identified as providing the least satis-
faction, coupled with the factors that are most correlated to job satisfaction, it
would appear that the Personal Feedback factor provides the greatest opportunity
for operational improvement. Future research should explore this factor in detail
to determine how hospitality operations could improve the satisfaction levels of
the Personal Feedback factor.
LIMITATIONS
The Downs and Hazen (1977) work in developing the CSQ occurred more than
20 years ago. Although subsequent work in numerous studies, discussed earlier,
generally supports the CSQ, some differences in factor structures have been iden-
tified. This is to be expected when dealing with a variety of subjects in a variety of
industries. The CSQ has survived as a useful model because of the general stabil-
ity of the original factor structure.
This study, in addition to the time factor, also dealt with the issues of homoge-
neity of management and ethnic background of the respondents. All six hotels are
managed by the same company and, due to the location of the hotels, all had a
large percentage of Hispanic employees. The findings in this research are limited
by these factors. Future research that builds from this work must acknowledge
that factor structures and results may be different based on the characteristics of
the sample. As Downs and Hazen (1977) suggest, and we agree, a confirmatory
factor analysis should be a prerequisite for any research using the proposed fac-
tors identified in this study.
APPENDIX
Listed below are several kinds of information often associated with a person’s job. Please
indicate the extent of your agreement with the following statements by placing the number
0-1-2-3-4-5-6-7-8-9-10 in the blank space provided. Let 0 represent strong disagreement, 5
represent neither agreement nor disagreement, and 10 represent strong agreement.
________ 4. I receive information about my progress in my job.
________ 5. I receive information about personnel.
________ 6. I receive information about company policies and goals.
________ 7. I receive information about how my job compares with others.
________ 8. I receive information about how I am being judged.
________ 9. I receive recognition of my efforts.
________ 10. I receive information about departmental policies and goals.
________ 11. I receive information about the requirements of my job.
________ 12. I receive information about government action affecting my company.
________ 13. I receive information about changes in the organization.
________ 14. I receive information on how problems in my job are being handled.
________ 15. I receive information about employee benefits and pay.
________ 16. I receive information about company profits and financial standing.
________ 17. I receive information about accomplishments and/or failures of the
company.
________ 18. Upper management knows and understands the problems faced by
employees.
________ 19. Company communication motivates and stimulates an enthusiasm for
meeting company goals.
________ 20. Upper management listens and pays attention to me.
________ 21. People in my organization have great ability as communicators.
________ 22. My supervisor offers guidance for solving job-related problems.
________ 23. Company’s communication makes me identify with it or feel a vital part
of it.
Answer the following section only if you are a MANAGER OR SUPERVISOR (Non-
managers, skip to No. 47).
________ 42. My employees are responsive to downward directive communication.
________ 43. My employees anticipate my needs for communication.
________ 44. I do not have a communication overload.
________ 45. My employees are receptive to evaluation, suggestions, and criticism.
________ 46. My employees feel responsible for initiating accurate upward
communication.
47. What is your age? ________ (years)
48. How long have you been in
your current position? ________ (years/months)
REFERENCES
Schneider, B., & Bowen, D. E. (1993). The service organization: Human resources man-
agement is crucial. Organizational Dynamics, 21(4), 39-52.
Sparks, B. (1994). Communicative aspects of the service encounter. Hospitality Research
Journal, 17(2), 39-50.
Varona, F. (1996). Relationship between communication satisfaction and organizational
commitment in three Guatemalan organizations. The Journal of Business Communica-
tion, 33(2), 111-140.
Vroom, V. H. (1964). Work and motivation. New York: John Wiley.
Zeithaml, V. A., Parasuraman, A., & Berry, L. L. (1990). Delivering quality service: Bal-
ancing customer perceptions and expectations. New York: Free Press.
Daniel J. Mount, Ph.D., is an assistant professor in the School of HRRM, The Pennsylva-
nia State University (227 Mateer Building, University Park, PA 16802-1307; e-mail:
dmount@psu.edu) and Ki-Joon Back is a Ph.D. candidate in School of HRRM, The Penn-
sylvania State University (1400 Martin St. #3101, State College, PA 16803; e-mail:
kxb261@psu.edu).