iScale: Studying Long-Term Experiences through Memory - Modeling Users' Experiences with Interactive Systems - page 105

Ta b l e 5 . 4 Number of experience reports judged for the three dimensions of richness, elicited
through the three different versions of iScale: constructive, value-account and no-graphing
(control).

Name Constr. VA Control

Contextual information

Y: 45 27 38

a. Event: Does the participant recall one or more discrete
events that lead to the realization of the reported experience?

N: 101 91 154

Y: 20 16 14

b. Temporal: Does the participant recall temporal information
about the reported experience?

N: 126 102 178

Y: 10 11 16

c. Expectation: Does the participant recall his/her expectations
about the reported experience?

N: 136 107 176

out short reports that did not contain sufficient text to form this judgment. 72% of
reports in the constructive condition were coupled, 81% in the value-account and
69% in the control condition. 52 coupled reports had incomplete time information
(in either session), leaving a total of 273 complete coupled reports.
For each reported experience participants estimated the moment in time (i.e. days,
weeks or months after the purchase of the product) at which the experience took
place (see figure 5.3a). We used formula 5.1 to compute the distance (or conver-
gence) between the recalled points in time of both sessions (for a justification of the
logarithmic transformation see the appendix section).

Δ = Abs ( log ( t 2 ) − log ( t 1 )) (5.1)

An analysis of variance with the computed temporal distance between experi-

ence reports from session 1 and session 2 as dependent variable and mode of recall
(constructive, value-account, control) and product quality (ease-of-use, innovative-
ness) as independent variables displayed significant main effects for mode of recall,
F(2, 267) = 5.42, p

Δ

.01, h p =0.04, and for product quality, F(1,267) = 4.66, p .05,

h p =0.02, but not for the interaction between mode of recall and product quality,
F(2,267) = 1.81, p=.2, h p =0.01. Post-hoc tests using the Bonferroni correction re-
vealed that participants in the constructive condition were significantly more consis-
tent in estimating the time that an experience took place as compared to the control
condition (p

< <

.01, Cohen's d=4.41). No significant differences were established be-
tween the two graphing conditions (p=.91, d=1.29), or between the value-account
and the control condition (p=0.1, d=3.01). Last, a significant difference was found in
the temporal consistency of reports for both product qualities, where reports of ease-
of-use were temporally more consistent than reports of innovativeness, t(271)=2.8,
p

<

< 0.01, Cohen's d=3.