Information Technology Reference
In-Depth Information
Table 2. Diffusion indexes correlated with benefit measurement
Std. Error of
Estimate
Index
Pearson's R
Adjusted R²
F-test
Sig.
t-test*
Volume
.446
.155
.919
4.447
.003
.000
Breadth
.426
.160
.916
8.221
.001
.000
Diversity
.546
.259
.861
7.656
.000
.000
Depth
.542
.274
.852
15.357
.000
.000
which follows that the results from the test do
not describe the situation in SMEs. However,
this was a test of the model, and our ambition is
to do further research which then could provide
additional knowledge when comparing results
from the perspective of organizational size. Ad-
ditionally, this limitation will make it possible
to generalize future research results to a higher
grade. Further, the underlying theory might not be
as promising as we had hoped for, since Massetti
and Zmud's (1996) dimensions might not work
in the context of ERP system.
for our analysis. Table 2 summarizes the results
of the four regression analysis. Returning to the
propositions, we can conclude that volume and
breadth received low support in this test, whereas
diversity and depth received support.
discussion And concLusion
The purpose of this paper was to develop and
test a model describing the connection between
diffusion of ERP and Perceived Net Benefits.
Diffusion of ERP was measured as the extent of
ERP systems' utilization from the four dimensions:
volume, breadth, diversity, and depth. These four
latent constructs were measured through two to
eight individual items. However, only two of the
four dimensions were found to be statistically
verified. One conclusion that can be drawn from
the fact that we did not get significant results on
the dimensions, volume and breadth, could be
that ERP systems historically are systems that are
used to a higher extent at the operational level, and
therefore could be seen as a commodity - necessary
to have but not providing any perceived benefits
from the respondents' views. Another potential
explanation is our interpretation of the underlying
model/theory, i.e., Massetti and Zmud's EDI us-
age measurement , which was either incorrectly
interpreted, or it could be that the model was not
applicable to the ERP context. A third possible ex-
planation is the operationalization of Massetti and
Zmud's four dimensions. Hence, the measurement
model needs to be further developed so that it more
resuLts
We used multivariate linear regression to mea-
sure the association of ERP utilization (volume,
breadth, diversity, and depth) and perceived net
benefits (operational benefits). Prior to the linear
regression analysis on the scores of the items in
the questionnaire, all items were standardized.
This was done since we used different scales
(percentage and Likert). The four dimensions
or latent constructs (volume, breadth, diversity,
and depth) of ERP utilization were tested against
the dependent variable ( net benefits , which was
measured through the item operational benefits).
For the actual analysis, we ran four separate linear
regression analysis. A more advanced statisti-
cal analysis could be considered, but since this
research was in its early stage, we explored the
underlying construct prior to testing the relation-
ships among constructs. We used SPSS version 17
 
Search WWH ::




Custom Search