Travel Reference
In-Depth Information
selected through systematic sampling. The rationale
for choosing this group of students is due to their
experiences as first time job seeker after graduate
from university. In addition they were also selected
because they have already working experiences dur-
ing their practical or internship training in tertiary
program. Data was collected via mail survey and
self-administered questionnaires to 760 respondents
however only 450 questionnaires were returned.
Hence, only 425 questionnaires that were found use-
ful and have been retained for further analysis.
percent (
N
= 130) held a diploma certificate while
28.7 percent (
N
= 122) were graduates from Tourism
Management program. It was found that all of the
respondents had internship training experience or
practical work experience in hospitality industry with
duration of training for 3-6 months. The respond-
ents' present status show that majority of gradu-
ates were employed with 76.2 percent (
N
= 324), 9.9
percent (
N
= 42) were unemployed and 13.9 percent
(
N
= 59) furthering their study to the higher level.
4.2
Structural equation of hypothesized
final model
Based on the modification index of CFA, the
measurement model of exogenous and endog-
enous and the final model as the examination of
the hypothesized model confirmed the constructs
of GE, CDL and WIL of the hypothesized paths.
In SEM, factor analysis and hypotheses are tested
in the same analysis. SEM techniques also provide
fuller information about the extent to which the
research model is supported by the data. Goodness
of fit indices for the 12 observed variables of CDL,
WIL and GEB shows that the reading is good if
it ranges from 0.188 to 0.941 for the significance
standardized regressions weight. Standard error
(SE) for each observation shows the goodness of
fit and low level reading from .014 to .029, and esti-
mate (Square Multiple Correlation) of observation
shows the contribution level to the latent variable
(.099 to 0.885). The standardized regression weight
between CDL and GE is 0.525, and between WIL
and GE is 0.296. The final model shows the model
explained in a substantial portion of the variance in
all the endogenous variables (square multiple corre-
lations) that indicates the two exogenous variables
(CDL and WIL) jointly explained 42.3 percent
variance in GE. Finally, from the Structural Model
the reading for GFI is at 0.90 (acceptable fit crite-
ria) and RMSEA is less than .08. The measurement
model has a good fit with the data based on assess-
ment criteria such as GFI and RMSEA (Bagozzi
& Yi, 1988). Table 1 summarizes the goodness of
3.2
Measures
A 22-items instrument used to measure career
development learning was adopted from the out-
come of career development learning, namely
the DOTS model as listed by Watts (2008). The
24-items used to measure work integrated learning
through internship training adapted from Muha-
mad, Yahya, Shahimi and Mahzan (2009), Dick-
erson& Kline (2008) and Singh & Dutta (2010).
Employability was measured using 39-items
adapted from Knight & York (2002) pertaining on
aspect of employability that graduates should pos-
sess to enhance the employability. All of the items
were measured by using the 5-point Likert Scale.
3.3
Data analyses
The demographic information was used to provide
an overview of respondents' profile. Principal fac-
tor analysis was performed to reduce the number of
factors or items from each variable. The final results
from this factor analysis were then used for further
investigation using Structural Equation Modelling
(SEM) with AMOS 20.0 program. It aims to find
the most optimal model or combination of the vari-
ables that fits well with the data on which it is built
and serves as a purposeful representation of the
reality from which the data has been extracted, and
provides a parsimonious explanation of the data
(Kline, 2011). In this study, the SEM technique was
used to identify the influence of career development
learning and work integrated learning to graduates'
employability upon graduation.
Table 1. Summary of the goodness fit of CDL, WIL,
GE and final model.
4 RESULTS
4.1
Profile of sample
Majority of the respondents 70.8 percent (
N
Model Fit Indicator CDL
WIL
GE
Final Model
301)
were female while the remaining 29.2 percent
(
N
=
(
x
2
)
10.322 7.203 128.489 182.060
DF
3
2
3.4
49
104) were male. The ages of respondents
involved were at a range from 22 and 26 years old.
Most of the respondents are 24 years old with 36.2
percent (
N
=
CMIN/DF
3.441 3.601
3.779
3.716
P
.016 0.027
.000
.000
GFI
0.990 0.992
0.945
0.937
=
154). Among of them, 69.4 percent
RMSEA
.076
.078
.080
.080
(
N
=
295) are degree holder graduates and only 30.6
Search WWH ::
Custom Search