Development the Measurement of Human Resource CAPABILITY IN
Hospitality IN UPPER NORTHEASTERN REGION OF THAILAND
Pana Dulalyaphut, UdonThani Rajabhat University, UdonThani, THAILAND
Asst.Prof.Subchat Untachai, UdonThani Rajabhat University, UdonThani,
THAILAND
ABSTRACT
The paper is designed to provide a quantitative measure of the human resource capability of
the hospitality in the upper Northeast of Thailand. The objective of this study is to examine the
validity and reliability of the measure of the four-factor model of the human resource capability
of the hospitality in the upper Northeast of Thailand.
The research mainly involves a survey design. It includes a pilot test using undergraduate
business students at UdonThani Rajabhat University for pretesting questionnaire items. In
addition, this investigation into skills, expertise, problem-solving, and adaptability capability
attributes necessitates uncovering variables of interest and this involves a large-scale field
study.
The data are collected via personal questionnaires from 209 samples. They include the
managers of hospitality firms in 3 provinces such as UdonThani, Nongkhai, BongKan.
Respondents are asked to rate, on a five-point Likert scale, their agreement or disagreement
on the human resource capability attributes. LISREL program is used for data analysis since
the proposed model is a simultaneous system of equations having latent constructs and
multiple indicators. Quantitative data are analyzed by the statistical techniques, namely
exploratory factor analysis and confirmatory factor analysis.
It is found from the study that the four - factor human resource capability, which consists of
skills, expertise, problem-solving, and adaptability capabilities of the hospitality in the upper
Northeast of Thailand, is empirically fit the data. The managerial implications are discussed.
Keywords: Human Resource Capability, Human Resource Management, Hospitality,
Confirmatory Factor Analysis,
1
INTRODUCTION
1. SMEs in Thailand, especially northeast of Thailand
2. Theory and concept of HR capability
2
X1,
X2,
X3,
Skills
1
12
X4,
X5,
X6,
13
expertise :
2
23
X7,
X8,
X9,
X10,
Solve:
3
14
24
34
X11,
X12
Adapt:
4
Figure 1. The HR capability Model
Based on this review of the literature relating to HR capability, twelve items of HR
capability were selected for inclusion in the survey questionnaire, that was used to
gather information for this article. Also, the authors proposed the following research
model illustrated in Figure 1 which offers a visual presentation of the four – factor HR
capability model.
This four HR capability model can be expressed as:
x   x  
Where x is a vector of 12 indicators,  is a vector of 4 HR capability constructs,  is a
124 matrix of pattern coefficients relating each indicator to its posited underlying
construct, and  is a vector of 12 indictor errors. The variance-covariance matrix for
the indicators denoted as , can be given as:
      
3
Where  is the 44 covariance matrix of HR capability constructs and   is the
1212 diagonal matrix of HR capability error variances.
OBJECTIVE AND HYPOTHESIS
The purpose of this study was to investigate the HR capability construct.
Specifically, the objective of this study was to examine the validity and
reliability of the measure of the four HR capability in the upper northeast of
Thailand. They were skills, expertise, problem-solving, and adaptability capabilities
in the upper northeast of Thailand.
H1) the HR capability consists
( 12 , 13 , 14 , 23 , 24 , 34
of skills, expertise, problem-solving, and adaptability capabilities
 0 ).
METHODOLOGY
The Sample and Data Collection
The research is based on a survey design. It includes a pilot test using undergraduate
students at Udon Thani Rajabhat University for pretesting questionnaire items. In
addition, this investigation into perceived cost of tourism, perceived benefit of
tourism, use of the tourism resource base, perceived state of the local economy, and
ecocentric attitude necessitates identifying variables of interest and this requires a
large-scale field study.
The sample was drawn from a list of all hospitality enterprises located at UdonThani
province (12,969 firms) and at Nongkhai province (24,201 firms), and Bongkan (1000
firms) (National Statistical Office, Ministry of Information and Communication
Technology, Thailand). From the initial list of 37,170 enterprises, a sample of 350
managers of the enterprises was quota selected. The data were collected via selfadministered questionnaires. Respondents were asked to rate, on a five-point Likert
scale their agreement or disagreement on the residents’ perceptions on ecotourism
dimensions. In June 2011- December 2011, 250 questionnaires were distributed to the
sample in upper northeast, Thailand.
Developing a better measure
The authors developed measurement items following the process that recommended
by Churchill (1979), and Gerbing and Anderson (1988). The first task was to generate
items, sample items and dimensions from researchers who have previously developed
the scale (Dyer et al., 2007; Jurowski and Gursoy, 2004; Ko and Stewart, 2002).
Second, the questionnaire items were submitted to three academic experts in the fields
of HR management for review. They were asked to review the survey for domain
representativeness, item specificity, clarity of construct, and readability (i.e. content
4
and face validity). Drawing on their inputs, some measurement items were eliminated
or reworded, and others were added. Third, the resultant survey instrument was
pretested with 30 undergraduate students in Thailand. They were asked to complete a
survey and indicate any ambiguity or other difficulties they experienced in responding
to the items. Their feedback and suggestions were used to modify the questionnaire.
These completed pilot test responses were analyzed with SPSS. An exploratory factor
analysis using Principal Component Extraction indicated that all items load on
expected factors (loadings range from 0.761 to 0.898). Construct reliability tests with
Cronbach's Alpha also yielded satisfactory results (range from 0.74 to 0.85). Finally,
12 items were verified with confirmatory factor analysis using LISREL 8.30. After
the iterative process of item refinement and purification, a battery of 73 items was
reduced to the final set of 12 items to measure the four constructs such as skills,
expertise, problem-solving, and adaptability capabilities.
Additionally the twelve
structured items (measured on a five-point scale) were rooted in agreement or
disagreement on the perceptions in ecotourism dimensions. This study has utilised the
instruments (see TABLE 2) to develop the HR capability model in upper northeast,
Thailand.
TABLE 2: HR capability Measures
Scale
Scale Items
skills,
X1. (Change Agent/Change leader)
X2. (Team Motivation)
X3. (Resource Allocation)
expertise,
X4. (Human Resource Management)
X5. (Personal Motivation)
X6. (General knowledge)
problem-solving,
X7. (Context Clarification)
X8. (Problem analysis)
X9. (Conflict Management)
X10. (Information sharing)
X11. (Flexibility)
X12. (Intelligence)
Adaptability
Validity
This study adopted the Gerbing and Anderson (1988) methodology to determine the
construct, and discriminant validity of the HR capability measures. To determine the
convergent and discriminant validity of the HR capability, measures were also
included in the questionnaire. These cover skills, expertise, problem-solving, and
adaptability capabilities. Discriminant validity is required when evaluating measures
5
(Churchill, 1979), especially when the measures are interrelated, as in the case of the
HR capability in the hospitality sector.
Analytical techniques
Before the data were analyzed, the questionnaires returned were reviewed to ensure
that appropriate information was being collected and defective questionnaires were
discarded. There were 209 completed questionnaires. The response rate of 84% was
high. The complete questionnaires were coded and the data keyed into the computer.
This paper mainly employed two statistical techniques to analyze the data. They were
exploratory factor analysis and confirmatory factor analysis (Bollen, 1989; Hulland et
al., 1996). At this time the LISREL was applied to the analysis process and a data
analyst was employed to supervise(Byrne, 1998).
TABLE 3: Means, Standard Deviations, Covariance Matrices
x1
x2
x3
x4
x5
x6
x7
x8
x9
x10
x11
x12
Mean
SD.
x1
0.55
0.40
0.37
0.34
0.29
0.27
0.35
0.32
0.26
0.23
0.30
0.31
4.05
0.74
x2
x3
x4
x5
x6
x7
x8
x9
x10
x11
x12
0.63
0.40
0.34
0.29
0.27
0.33
0.34
0.28
0.25
0.34
0.33
4.07
0.79
0.60
0.36
0.28
0.30
0.30
0.27
.27
.23
.32
.33
4.10
0.77
.55
.36
.25
.30
.31
.26
.22
.34
.31
4.06
0.74
.55
.27
.24
.27
.24
.21
.30
.33
4.08
0.74
.55
.36
.28
.26
.23
.30
.29
4.15
0.74
.60
.39
.29
.26
.32
.34
4.17
0.78
.65
.41
.29
.37
.37
4.09
0.80
.61
.30
.32
.31
4.12
0.48
.51
.28
.34
4.14
0.72
.60
.43 .73
4.14 4.10
0.77 0.85
RESULTS
Hypothesis Testing
Assessing fit between model and data
The analysis began with calculation of the mean and standard deviation for each
unweighted, interval scale. Covariance between each scale is reported in Table 3. The
overall adequacy of the proposed theoretical framework was examined using LISREL
8.30 causal modeling procedures (Joreskog and Sorbom, 1996), and the maximum
likelihood method of estimation and the two-stage testing process were adopted. A
substantial portion of the variance in the residents’ perception on ecotourism has been
explained by the model. The results are shown in Table 4. The model is a close fit to
the data at 2 (48) value of 112.05 (P<0.00). However, the ratio of Chi-square and
degree of freedom is 2.33 (112.05 /48), GFI of 0.92, AGFI of 0.87, CFI of 0.98 and
RMSEA of 0.08. Therefore, the residents’ perception on ecotourism model can be
acceptable (Bentler, 1990; Bentler and Bonett, 1980) (see Fig. 2).
6
TABLE 4: Properties of the CFA for HR capability
Construct
skills,
expertise,
problemsolving,
Adaptability
capabilities
Items
X1
X2
X3
X4
X5
X6
X7
X8
X9
X10
X11
X12
Standardize
d loadings
t-value
CR
AVE
Cronbach
’s Alpha
.84
.82
.79
.80
.73
.62
.70
.79
.79
.70
.84
.77
13.50*
12.96*
11.08*
9.13*
10.31*
10.41*
9.25*
11.91
.70
.66
.85
.60
.53
.74
.62
.55
.83
.69
.64
.78
*Indicates significance at p<.01 level
Assessing reliability and validity of constructs
Before testing the hypotheses, evaluated the psychometric properties of the
measurement scales through confirmatory factor analysis using LISREL (Anderson
and Gerbing, 1988). The composite reliability(CR), variance extracted estimates
(AVE), convergent validity, and discriminant were examined.
Composite reliability reflects the internal consistency of the indicators in measuring a
given factor (Fornell and Larcke, 1981). The composite reliability values for each of
the HR capability dimensions is shown in Table 4 which reveals that the composite
reliability score for each dimension is satisfying (0.70, 0.60, 0.62, 0.69). In addition,
the Cronbach’s alpha values for each of the HR capability dimensions are shown in
Table 3, which in each case is greater than 0.60(Bagozzi, 1988). In addition, the result
was that the variance extracted estimates construct are all a greater than .50 (.66, .53,
.55, .64).
Besides the reliability test, convergent validity was demonstrated when different
instruments were used to measure the same construct, and scores from these different
instruments are strongly correlated. The convergent validity can be assessed by
reviewing the t-test for the factor loadings (greater than twice their standard error)
(Anderson and Gerbing, 1988). The t-test for each indicator loading is shown in Table
4. In the result of this analysis the construct demonstrates a high convergent validity
because all t-values are significant at the .01 level.
7
In addition, the confidence interval test to assess the discriminant validity between the
four factors involves calculating a confidence interval of plus or minus two standard
errors around the correlation between these factors, and determines whether this
interval includes 1.0. If it does not include 1.0, discriminant validity is demonstrated
(Anderson and Gerbing, 1988). Table 5 shows the values of interval between 2
factors. They were 0.81, 0.97; 0.77, 0.93; 0.77, 0.93; 0.74, 0.90; 0.79, 0.99; and 0.70,
0.86. That is to say that discriminant validity for the HR capability scale is
significantly supported because all range excludes the value 1.0.
The correlation coefficients between each pair of factors for the HR capability scale; e.g.
skills, expertise, problem-solving, and adaptability capabilities, are significant. They are
12= 0.33 (t =7.68, p<0.05), 13= 0.27 (t=6.90, p<0.05), 14= 0.32 (t=7.25, p<0.05),
23= 0.27 (t=6.87, p<0.05), 24= 0.35 (t=7.63, p<0.05), 34= 0.29 (t=7.02, p<0.05), .
Therefore, the hypothesis testing is supported.
TABLE 5: Test of Discriminant Validity for AVE and Confidence Interval
Skills
expertise
problem-solving
adaptability
.70
.79
.67
.60
expertise
(.81, .97)
.60
.72
.79
problem-solving
(.74, .90)
(.77, .93)
.62
.73
adaptability
(.70, .86)
(.79, .99)
(.77, .93)
.69
Notes: On the diagonal the AVE of each factor is shown. In the upper part, the square of the
correlation between each pair of factors is detailed and the confidence interval for every pair of
factors is collected in the lower part.
Skills
8
Figure 2. Standardized Solution of HR capability model
DISCUSSION
REFERENCES:
Anderson, J. C., & Gerbing, D. W., ”Structural equation modeling in practice: a review and
recommended two-step approach,” Psychological Bulletin, Vol. CIII (3), 1988, 411-423.
________. “An approach for confirmatory measurement and structural equation modeling of
organizational properties”, Management Science, Vol. XXXIII (4), 1987, 525-541.
Bagozzi, R. P. & Yi, Y., “On the evaluation of structural equation models,” Journal of Academy of
Marketing Science, Vol. XVI, 1988, 74-94.
Bentler, P. M. and Bonett, D. G. ,”Significant tests and Goodness-of-Fit in the analysis of covariance
structures,” Psychological Bulletin, Vol. LXXXVIII(3), 1980, 588-606.
________., ”Comparative Fit Indexes in structural models,” Psychological Bulletin, Vol. CVII (2), 1990,
238-246.
Bollen, K. A. (1989). Structural Equations with Latent Variables. NY: John Wiley & Sons, Inc.
Byrne, B. M. (1998). Structural Equation Modeling with LISREL, PRELIS, SIMPLIS: Basic concepts,
applications, and programming, Mahwah, NJ.
Churchill, G. A., “ A Paradigm for Developing Better Measures of Marketing Constructs,” Journal of
Marketing, Vol.16(February), 1979, 64-73.
Fornell, C., & Larcker, D. F., “Evaluating Structural Equation Models with Unobservable Variables and
Measurement Error,” Journal of Marketing Research, Vol. XVIII(February), 1981, 39-50.
Gerbing, D. W., and Anderson, J. C., “An updated paradigm for scale development incorporating
unidimensionality and its assessment,” Journal of Marketing Research, Vol. XXV (May), 1988,
186-192.
Hulland, J., Chow, Y. H., & Lam, S., “Use of causal models in marketing research: a review,”
International Journal of Research in Marketing, Vol. XIII, 1996, 181-197.
9