Agneta & Frida - Narrative Intelligence - page 250

While the subjects' description of the system would reflect the more con-
scious aspects of the experience, we employed another method to capture more
deep-level dimensions. Inspired by Maglio and Matlock (1999) and Lakoff and
Johnson (1980, 1999), we performed a metaphor analysis of the interviews.
From Maglio & Matlock's study, we knew that web browsing is often per-
ceived as a spatial activity: the user is viewed as an agent moving through the
space of sites and web pages. Maglio and Matlock found this by examining
the metaphors used when subjects described their surfing through web pages:
'I browse/surf the web'; 'I go to pages'; 'I enter/leave pages'; 'pages contain
information'; 'the web is an information space in which I look for things'.
We decided to follow the method used by Maglio and Matlock, focusing
on narrative versus spatial verbs and adverbs in the interviews. Spatial verbs
and adverbs were characterized by movement (e.g. 'going through'). As for
narrativity, we looked for words containing temporal dimensions (e.g. '. . .and
then... ) or intentional/psychological words (e.g. 'giving up'; 'bored'; 'anx-
ious'). In contrast to spatial experiences, we hypothesized, narratives are tem-
poral and causal chains of events and this would have to be reflected in the use
of verbs and adverbs.
Finally, we measured disturbance and recall . If the user was able to integrate
the narrative of Agneta and Frida with the web content, we hypothesized, sub-
jects would be less disturbed by the two characters than by a case in which the
Agneta & Frida story ran 'in parallel' to the web content. In the latter case, the
comments and activities of the characters would be experienced as intrusive. As
for recall, we assumed that the emotional reactions caused by the remarks from
Agneta & Frida - e.g. laughs, frustration, moral judgment and agreeableness -
would enhance the recall of the information remarked upon. We assumed that
Agneta and Frida would encourage the user to construct a narrative context
and associative links between information in the site, which would improve
memory. Thus, we expected the Agneta and Frida subjects to perform better
on a post-usage recall test, than would subjects without Agneta & Frida.
Subjects
The 38 subjects were recruited to be between 20 and 30 years. Eighteen subjects
used Agneta & Frida (the 'withA&F' group) and 20 subjects explored the web
sites without the characters present (the 'withoutA&F' group).
The subjects of the withA&F group were in the range 19 < 26.2 < 41 years
old (seven women and eleven men). Ten had a technical background; the rest
had other professions. All but one had a university degree. The subjects in