Information Technology Reference
In-Depth Information
7.3.3
Empirical Knowledge Bases for Forming Design Goals
Designers form beliefs. In the absence of empirical insights, designers base these
beliefs on prior experiences and their own intuition. Empirical knowledge derived
from similar artifacts in similar contexts may not only make these beliefs grounded
on actual evidence, but may also inspire the process of forming beliefs. Computa-
tional infrastructures such as the ones described above may provide rich insights
into the relationships between design parameters, contextual parameters, and their
psychological consequences (e.g. product quality perceptions).
7.3.4
A New Basis for User Insights?
The computational infrastructure proposed above assumes the use of structured in-
terview techniques such triading, laddering and pyramiding; these share the ben-
efit of imposing a set of relations between the constructs being elicited and thus
are more “computational-friendly”. The iScale tool lead to a new challenge rooted
in providing access to immense, unstructured qualitative data. Chapter 6 proposed
a computational approach for the content analysis of unstructured qualitative data
such as in the case of experience narratives. One may however note that content
analysis is limited in its scope as it looks only for one type of relation, that of
similarity/dissimilarity between different data instances (e.g. narratives). The ques-
tion then becomes: how can computational tools assists researchers in identifying a
richer set of relationships such as causal effects in unstructured qualitative data? Can
this procedure be partly automated, e.g. uncovering causal relationships through the
relative position of concepts within the unit of analysis? If not, how can we assist the
researcher in transforming qualitative data to a computational format (i.e. a graph
network) that would enable him in rapid hypothesis testing, e.g. counterfactual anal-
ysis (King et al., 1994) within her own dataset as well as across different datasets?
Next, given the access to computational tools that partly automate the analysis of
rich qualitative data and support the identification of emerging patterns, we foresee
a new basis for user insights, one derived from internet forums, complaint centers,
emails and other sources of public statements. These may lead to user insights re-
lating to product success in the market but also to emerging trends and unidentified
needs in the marketplace. With this work, through the development of the iScale
tool and the Computational Content Analysis method, we aimed at making a first
step in inquiring into richer sources of user insight data.
