The mediating effect of attitude between customer responsiveness and actual usage of “Touch ‘n Go” card in foodservice outlets - Theory and Practice in Hospitality and Tourism Research - page 357

560 were distributed to the customers in the partici-
pated restaurants. Of the total number of respond-
ents, resulting in an effective sample of 400 usable
completed surveys and thus met the preliminary
screening requirements, meanwhile representing
for an overall response rate of 71.4 percent. Prior
to the analysis, the questionnaires were screened for
missing data, response patterns that indicated low
discrimination ability and logic inconsistencies. A
total of 160 questionnaires were removed from the
data analysis since the questionnaires were incom-
plete and not suitable for use (Tabachnick & Fidell,
2007) leaving a sample of 400 respondents for the
analysis. The data were analyzed using IBM SPSS
Statistics version 21 for Windows together with
INDIRECT macro to assess the mediating effect
(Preacher & Hayes, 2008). The descriptive statistics
performed in support of background information
of the customers.

Table 1. Distribution of respondents according to
background information.

4 RESULTS AND DISCUSSION

The demographic characteristics of the respond-
ents are displayed in Table 1. All information is
presented in actual figures and percentages to
facilitate interpretation.
The majority of the respondents were Malays
(76%), followed by Indians (8.5%), Chinese (5.5%)
and others (10%). More than half of the respond-
ents were females (61%), while the remaining (39%)
were males. As for age, the majority of the respond-
ents were between 19 to 30 years old (66%). Merely
21.5 percent of the respondents had a first-degree
qualification. The other 78.5 percent had Diploma,
Certificate or other qualifications. The majority of
the respondents work in other various occupations
(39.5%). A total of 19 percent among the non-execu-
tive, executive (18.5%), professional (16.5%), home-
maker (4%), senior management (1.5%), and retired
(1%). A majority of the respondents earned between
RM1001 to RM2000 per month (56%) and followed
by RM2001 to RM3000 per month (19.8%).
In measuring the internal consistency of Cus-
tomer Responsiveness dimensions (Perceived
Usefulness, Perceived Ease of Use, Perceived
Trust) and Attitude, Cronbach's alpha were used.
Table 2 indicates that all dimensions in Customer
Responsiveness had good internal consistency
(0.835-0.880) while Attitude had excellent internal
consistency of 0.913.
Table 3 summarized the bootstrap estimates
based on 5,000 bootstrap samples. The mean indi-
rect effects from the bootstrap analysis were posi-
tive for all customer responsiveness dimensions and
significant (PU: a x b

Table 2. Reliability for Customer Responsiveness dimen-
sions and Attitude.

to 0.26, PEOU: .03 to 0.30, PT:.05, 0.47). In the
indirect path, the respondents who indicated high
agreement level of Perceived Usefulness, Perceived
Ease of Use and Perceived Trust, were more likely
to have better attitude (PU: a =0.21, PEOU: a =0.23,
PT: a =0.45); while holding constant for Customer
Responsiveness , the respondents who had better
attitude, will also have a high level of actual usage
( b =0.57). Based on the significant mean indirect

= 0.12, PEOU: a x b = 0.13, PT:

a x b 0.26), with a 95% bias corrected and accel-

erated confidence interval excluding zero (PU: .04

=