User Experience Over Time - Modeling Users' Experiences with Interactive Systems - page 61

D ij =1-—R ij — between the individual attributes i and j was derived from the cor-
relation R ij between the ratings by different participants on the two attributes. The
obtained distances were subsequently visualized in three dimensions (Stress value
S=0.19 1st week; S=0.15 4thweek) using the Multidimensional Scaling tool XGms
(Martens, 2003). Hierarchical clustering (with minimum variance) was applied to
the predicted distances in the 3-dimensional space. Figure 4.3 illustrates a 2D pro-
jection of the 3-dimensional visualization of the distances between the quality at-
tributes. The obtained clusters are denoted by the connecting lines. The left figure
reflects the users' ratings during the first week of use while the right figure reflects
the users' ratings after four weeks of use. All in all, clusters derived from the mea-
surement in the first week reflect the assumed underlying constructs, with a close
knit groups of mainly pragmatic and hedonic stimulation attributes and a looser rest
of hedonic identification attributes. After four weeks, the perceptions seem much
more differentiated, and relationships among attributes lost some of their strength.

Ta b l e 4 . 1 Bipolar attributes measuring pragmatic quality , stimulation and identification
(Hassenzahl, 2004)

PragmaticQuality Stimulation Identification

Technical-Human Typical-Original Isolating-Integrating

Complicated-Simple Standard-Creative Amateurish-Professional

Impractical-Practical Cautious-Courageous Gaudy-Classy

Cumbersome-Direct Conservative-

Innovative

Cheap-Valuable

Unpredictable-
Predictable

Lame-Exciting Noninclusive-Inclusive

Confusing-Clear Easy-Challenging Takes me distant from people -

Brings me closer to people

Unruly-Manageable Commonplace-New Unpresentable-Presentable

During the first week, Beauty judgments relate mostly to attributes reflecting the
quality of stimulation (i.e. original, creative, new, innovative) and to one attribute
reflecting identification (i.e. classy). This close relationship between stimulation and
beauty seems to have disappeared after four weeks of use; beauty now seems to be a
disjoint judgment only related to one identification (i.e. classy) and one stimulation
(i.e. challenging) attribute. While the relation between “classy” and beauty remained
over the period of four weeks, stimulation seemed to influence beauty judgments
mostly during the first experiences. Surprisingly, stimulation seemed to be a more
important determinant of beauty than identification in the first experiences.
A similar pattern can be observed for judgments of Goodness. During the first
week judgments of goodness relate mostly to pragmatic quality attributes (i.e. prac-
tical, direct, manageable, predictable, clear) and to one attribute reflecting identi-
fication (i.e. presentable). After four weeks of use goodness appears to be related
mostly to identification (i.e. professional, inclusive, valuable, integrating, brings me