www.sharexxx.net - free books & magazines
www.sharexxx.net - free books & magazines
www.sharexxx.net - free books & magazines
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
196 Mullany and Lay<br />
Of significance is the question of whether or not complaints were distinguished<br />
in a consistent manner. With this in mind, notes were carefully taken at each<br />
interview of any unusual occurrences and responses, and how they were handled.<br />
Similar action was thus ensured where similar circumstances arose subsequently.<br />
For example, experience soon showed that responses to seven-point scales<br />
required special handling, because several of the respondents could not remember<br />
all the options at once. Even though the respondent was shown the form, he often<br />
required help in making his assessment. This assistance was given in the form of a<br />
two-tiered approach. First, the respondent was asked to make a crude assessment<br />
out of options 2, 4 and 6, and then he was asked to refine his choice by making a<br />
selection within one of his original choice. For example, if he first chose option 6,<br />
he would then be asked to make a final assessment out of 5, 6 or 7.<br />
The criticism that points of complaint were recorded to correlate with KAI<br />
scores and/or differences was counteracted in the research design as follows: the<br />
user R-score was measured close to the beginning of the interview, before the user’s<br />
KAI score was measured, and before it became possible to make a reliable guess<br />
of the user’s KAI score. Other questions on the System Satisfaction Schedule were<br />
not asked until after the R-score had been measured in the case of a user, once again<br />
to avoid guesses as to the user’s cognitive style.<br />
FINDINGS AND DISCUSSION<br />
In this study, thirty-four systems were researched, hence the data is divided<br />
into several univariate samples of thirty-four (that is, sample-size n = 34 in all cases).<br />
The levels of significance employed in this study were based on the opinions<br />
found in human science and statistical literature. These opinions are summarized in<br />
Table 1. Based on the opinions of respected experts as recorded in Table 1, the<br />
qualitative ratings listed in Table 2 were assumed for this study. The data was then<br />
stratified into those univariate samples identified in Table 3.<br />
The mean and standard deviation was calculated for each sample, which was<br />
then tested for goodness-of-fit to the normal distribution.<br />
The Randomness of the Data Samples<br />
The means for the analyst and user KAI score samples were respectively,<br />
102.9 and 101.6, while their respective standard deviations are 12.55 and 14.09.<br />
Kirton’s British sample of KAI scores for 562 persons exhibited a normal<br />
distribution with mean 95 and standard deviation 18 (Kirton, 1987). In studies cited<br />
by Kirton, in which the KAI scores for various occupational groups were<br />
Copyright © 2003, Idea Group Inc. Copying or distributing in print or electronic forms without written<br />
permission of Idea Group Inc. is prohibited.