iScale: Studying Long-Term Experiences through Memory - Modeling Users' Experiences with Interactive Systems - page 95

Table 2, line Overall, shows the distribution of the four different types of curves.
The majority of segments (44 of 74, 60%) were categorized as linear. Only 5% (4
of 74) of segments were non-linear. Of those, only a single report was associated
with two or more linear segments with different slopes (cf. figure 5.4a segment
2, figure 5.4b segment 6, figure 5.4d, segment 1). In addition, only 4 of 74 (5%)
instances of discontinuity were observed in the graphs. Thus, while in some cases
users are inclined to draw non-linear curves, the majority of curves were linear.
This allowed us to focus iScale on linear graphing, thereby reducing a number of
potential problems with handling complex, non-linear types of curves in an online
tool.

Ta b l e 5 . 2 Relationship between graphing and reporting in free-hand graphing. Types of
graphing: C=Constant, L=Linear, NL=Non-Linear and D=Discontinuous

Type of graph

Type of report C L NL D

Discrete experience 3 30 2 2 37 (50%)

Overall evaluation without mo-
tivation

17 4 1 2 24 (32%)

Overall evaluation, motivated 2 10 1 0 13 (18%)

by experience

Overall 22 (30%) 44 (60%) 4 (5%) 4 (5%) 74

To get an idea of how the graphs relate to reported detailed experiences, the ob-
tained verbal reports were classified into three broad categories. One category is the
recall of a distinct experience : an experience that relates to a particular event with
beginning and end. According reports were indicative of the constructive mode:
recalling contextual information from a specific experience was followed by the
reconstruction of the value judgment from the recalled facts. For example: “The
reason I got this device was to develop applications for it. [the company] has a
special program for educational institutions to which provides free licenses for de-
velopment. But when we contacted them, they even questioned the existence of our
institution... this should have happened around here [points to the curve]” .Such
distinct reports provided one or more contextual cues about the past experience,
such as temporal information (i.e. when the event took place), changes in the con-
text of use (e.g. “then I went on vacation...”), information related to the participant's
social environment (e.g. “a friend of mine was very positive...”), etc. They consti-
tuted the most dominant type of reporting (37 of 74, 50%).
Other reports provided no contextual information about a recalled experience, but
instead, the participant reported an overall evaluation without further motivation:
“after that, [my opinion] is about constant, it hasn't changed lately” . Such reports
are typical for a pure value-account mode of recall: recalling an overall evaluation
of a specific experience or period, while failing to recall contextual cues or facts
about an exact experience (24 of 74, 32%).