Personal Attribute Judgments - Modeling Users' Experiences with Interactive Systems - page 17

Chapter 2
Personal Attribute Judgments

Abstract. Traditional approaches to measuring users' responses to artifacts lie in
the a-priori definition of the measures by the researchers. This chapter highlights
the limitations of such approaches that employ standardized psychometric scales
and introduces personal attributes judgments. It introduces attribute elicitation tech-
niques and in particular, the Repertory Grid Technique (RGT). It argues that the
true value of RGT is in quantifying rich qualitative insights and highlights the lim-
itations of relevant statistical techniques that are typically employed in the analysis
of repertory grid data. An initial Multi-Dimensional Scaling (MDS) procedure that
aims at identifying diverse views in Repertory Grid data is proposed. The procedure
identifies distinct user groups in a sample population and derives a two-dimensional
view for each respective user group. The technique is presented through a case study
where users' views on a set of product concepts were contrasted to the ones of de-
signers. The technique revealed differences not only between users and designers
but also between designers of different professional background and role in the de-
sign team.

2.1 Introduction

Reductionist approaches to user experience evaluation are grounded on the assump-
tion that people may summarize experiences in overall evaluative judgments (Has-
senzahl, 2008). When a product is associated with a certain experience, the value of
the experience will be partially attributed to the product. Such evaluations may be
substantive (Hassenzahl, 2004), e.g. perceptions of a given product quality such as
usability, or verdictive , e.g. overall evaluations of goodness, appeal, or beauty.
These are all latent constructs, in that they cannot be directly measured but in-
stead, they are estimated through a number of different measures. For instance, the
Unified Theory of Acceptance and Use of Technology (Venkatesh et al., 2003) mea-
sures a latent construct termed performance expectancy through a set of four items
such as “I would find the system useful in my job” and “Using the system increases
my productivity”. These items are assumed to be measuring different facets of the