Travel Reference
In-Depth Information
3 METHODOLOGY
As illustrated in Table 1, when CSR activities were
regressed on guest retention, the five CSR activi-
ties were able to explain 38.0 percent ( R 2
The survey questionnaire was personally distrib-
uted to hotel guests who have had experiences
staying at twenty four five-star hotels in Kuala
Lumpur. From the 600 questionnaires distributed,
a total of 283 questionnaires (47.2%) were com-
pleted, coded and analyzed. Regression analysis
was used to examine the relationship between vari-
ables selected in this study. Guidelines on media-
tion by Baron and Kenny (1986) were used to test
the mediating effect of brand image. The Cron-
bach's Alpha for the study variables ranged from
0.61 to 0.92.
=
0.38,
F-Change
.01) of the variance in guest
retention (Model 1). From the five CSR activities,
community involvement in quality of life (
=
34.02, p
<
β =
0.39,
p
<
.01) and environmental practices (
β =
.24,
p
.01) significantly contributed to the prediction
of guest retention, followed by involvement in com-
munity project (
<
.05).
In Model 2, when CSR activities were regressed
on brand image, CSR activities were able to explain
53.0% ( R 2
β =
.15, p
<
.01) of
the variance in hotel brand image. Involvement
in community project had the most impact on the
prediction of hotel brand image (
=
.53, F-Change
=
63.24, p
<
β =
0.43, p
<
.01),
4 RESULTS
environmental practices (
β =
0.20, p
<
.01), followed
by sustainable development (
β =
0.13, p
<
.05) and
Table 1 shows the results of the regression analysis
of CSR activities on guest retention.
community involvement in quality of life (
β =
0.12,
p
.05).
The procedures as suggested by Baron and
Kenny (1986) were followed in order to examine the
mediating roles of hotel brand image. According to
Baron and Kenny (1986), the following conditions
must be present for mediation effects: (1) the IV
(CSR activities) must significantly affect the medi-
ator (Brand Image), (2) the IV must significantly
affect the DV (Guest Retention), (3) the mediator
must significantly affect the DV, and (4) the effect
of the IV on the DV shrinks upon the addition of
the mediator to the model. As shown in Table 1,
only environmental practices, involvement in com-
munity project, and involvement in quality of life
able to meet the conditions for mediation. Table 2
provides the summary of the regression analyses to
ascertain the mediation effects of brand image on
the relationship between CSR activities and guest
retention.
The results from Table 2 showed that hotel brand
image was significantly and positively predicted
guest retention (
<
Table 1. Summary of hierarchical regression of CSR
activities, and Brand Image (BI) on Guest Retention
(GR).
Model 1
Std. Β GR
Model 2
Std. Β BI
Predictors
CSR to Environments:
• Sustainable development
.03
.13*
• Environmental practices
.24**
.20**
• Energy supply
.02
.60
CSR to Community:
• Involvement in community project
.15*
.43**
• Involvement in quality of life
.39**
.12*
R 2
.38
.53
Adj. R 2
.37
.53
.01). On the mediat-
ing effect, the effect of environmental practices on
guest retention had a decreasing
β =
0.54, p
<
R 2 Change
.38
.53
F-Change
34.02**
63.24**
β
-value (
β =
0.17,
p
.01) in the presence of the hotel brand image,
implying partial mediation. Similarly, the effect
<
Note : N
=
283, * p .
<
05; ** p .
<
01, Five-point Likert Scale,
GR
=
Guest Retention, BI
=
Brand Image.
Table 2.
Impact of mediated regression.
Predictors
Model 1 Std. Β GR
Model 2 Std. Β BI
Model 3 Std. Β GR
CSR Activities:
• Environmental practices
.24**
.20**
.17*
• Involvement in community project
.15*
.43**
.02
• Involvement in quality of life
.39**
.12*
.35**
Brand Image
.54**
Note :. N
=
283, * p
<
.05; ** p
<
.01, Five-point Likert Scale, GR
=
Guest Retention, BI
=
Brand Image.
Search WWH ::




Custom Search