Information Technology Reference
In-Depth Information
0,40
0,40
0,30
0,30
0,20
0,20
0,10
0,10
0,00
0,00
ease-of-use
innovativeness
constructive
value-account
control
Quality
Mode of recall
Fig. 5.8 Mean temporal inconsistency, together with the 95% confidence interval, between
two recalls of the same experience, (left) when using the constructive, the value-account, and
the no-graphing (control) version of iScale, (right) when reporting on perceived ease-of-use
and perceived innovativeness.
5.4.2.4
Test-Retest Consistency of Graphs
Figure 5.9 displays example graphs by two participants in two respective sessions
with the constructive iScale and two participants with the value-account iScale. The
area between the two graphs is a simple measure for the inconsistency in partici-
pants' graphs. It was calculated through sampling the graphs in 100 steps.
An analysis of variance with this area measure as dependent variable and mode
of recall (constructive, value-account) and product quality (ease-of-use, innovative-
ness) being reported as independent variables revealed a significant effect for mode
of recall, F(1, 44) = 6.75, p
.05, h p =0.13, but not for product quality, F(1, 44) =
.481, p=.49, h p =0.01, or for the interaction between mode of recall and product
quality, F(1,44) = 0.073, p=.8, h p =0.002. Graphs elicited through the constructive
iScale tool were more consistent (M area =30.1, SD=17.4) than ones elicited through
the value-account iScale (M area =52.2, SD=36.7, t(46)=2.7, p
<
<
.05, cohen's d=0.8)
5.4.3
Discussion
As expected, the constructive iScale tool lead to a significant increase in the number
of elicited reports (with a large effect size, d=1, Cohen, 1992), but also the richness,
of elicited reports, as compared to the control condition that involved no graphing.
When using the constructive iScale, participants elicited approximately 50% more
experience reports, and were more likely to recall cues referring to a discrete event,
or ones referring to temporal information, such as when the experience took place.
Contrary to our expectations, the value-account iScale did not result in similar bene-
fits. These findings support the idea that graphing, through imposing a chronological
order in the recall process (Anderson and Conway, 1993), supports the reconstruc-
tion of the context in which experiences took place, thus forming stronger temporal
and semantic links across the distinct experiences (Kahneman et al., 2004).
Search WWH ::




Custom Search