Information Technology Reference
In-Depth Information
The lowest ranked risks of Web development
were inefficient use of personnel time, followed by
solving the wrong problem and lastly redundant
development effort. However, none of these risks
was discounted, with averages that indicate that
the majority of respondents recognised them as
risks of relative importance. Not one respondent
rated inefficient use of personnel time as 'not
important'; five (9.4%) rated solving the wrong
problem as 'not important' and one (1.9%) rated
redundant development effort as 'not important'.
The participants considered solving the wrong
problem and redundant development effort to
be greater risks when undertaking spreadsheet
development (t(52) = 2.06, p = 0.044; t(52) =
2.55, p = 0.014). This may reflect an increased
sophistication of end-user developed spreadsheet
applications compared to end-user developed Web
applications.
It was interesting to note that the average
importance of each risk was lower for the group
who had previously developed Web pages than for
the group who had not, although the differences
were only significant for four risks: inability to
identify correct and complete requirements (t(52)
= 2.36, p = 0.022), use of private systems when
organizational systems would be more appropriate
(t(52) = 2.19, p = 0.033), solving the wrong problem
(t(51) = 2.54, p = 0.014), and redundant develop-
ment effort (t(52) = 2.70, p = 0.009). A reason for
this difference could be that the development
process has given them insight that allows them
to discount the risks; however, this seems unlikely
given the prevalence of problems with end-user
developed applications. It would seem more likely
that the satisfaction they derive from their own
Web development allows an overshadowing of
the perceptions of risks. This should be explored
further in future studies.
Future research should also differentiate
between different types of Web applications
that might have different risks and benefits. For
example, the risks associated with end-user de-
veloped Web pages that merely display informa-
tion could be considered substantially less than
those associated with applications that process
information.
approaches to reducing the risks
of end user web
development
The approaches to reducing the risks of end-user
Web page development are ranked by perceived
importance in Table 8. This table also includes
the average importance of each of the approaches
reported for the predominantly information tech-
nology staff in the Nelson and Todd (1999) study.
The importance of each approach as perceived
by the end-user developers in the current study
is compared with the average obtained in Nelson
and Todd's study using one sample t-tests and the
results are also presented in Table 8. All of the
approaches were rated fairly high by the end-user
developers with averages above the midpoint of the
scale. The highest ranked approach was training.
As discussed above, previous studies have found
that end-user developers receive very little train-
ing and what they do get tends to be self-training
rather than formal training (Chan & Storey, 1996;
Hall, 1996). The results in this study regarding
training for Web development are consistent with
other forms of end-user development such as
spreadsheet development. The acknowledgment
of the importance of training is quite interest-
ing, as despite having received little training
themselves, the respondents considered training
to be the most important approach to reducing
the risks of end-user Web development. Nelson
(1991) suggested that training is perhaps the most
effective tool for minimising the risks associated
with end-user development and the results of this
study suggest that end users agree.
Policies for data management were considered
to be the second most important approach. This
was unexpected because end-user developers have
traditionally been dissatisfied with approaches
to the management of end-user computing that
Search WWH ::
Custom Search