Personal Attribute Judgments - Modeling Users' Experiences with Interactive Systems - page 20

likely to have employed shallow processing (Sanford et al., 2006), i.e. responding
to surface features of the language rather than attaching personal relevance to the
question.
An alternative approach to posing predefined questionnaires to participants lies
in a combination of structured interviewing, that aims at eliciting the attributes that
are personally meaningful for each individual, with a subsequent rating process per-
formed on the attributes that were elicited during the interview. Many different in-
terview approaches have been proposed in the fields of Constructivist and Economic
Psychology. For instance, Free Elicitation, rooted in theories of Spreading Activa-
tion (Collins and Loftus, 1975), probes the participants with a stimulus and asks
them to rapidly express words that come to mind. The Repertory Grid Technique
(RGT), rooted in Kelly's Theory of Personal Constructs (Kelly, 1955), provides
three alternatives to the participants and asks them to define dimensions in which
the three products are meaningfully differentiated. The Multiple Sorting Procedure,
rooted in Facet Theory (see Al-Azzawi et al., 2007), asks the participant to sort
products in a number of piles, and only later on define a label for each pile. Com-
paring the different techniques is not the focus of this work; see (Bech-Larsen and
Nielsen, 1999; Breivik and Supphellen, 2003; Steenkamp and Van Trijp, 1997; van
Kleef et al., 2005) for more information on this. While the analysis procedures that
have been developed in the context of this work are grounded on Repertory Grid
data, they may also be well applied to data derived from any of the other attribute
elicitation techniques.

2.2 The Repertory Grid Technique

The RGT is one of the oldest and most popular attribute elicitation techniques. It
originates from Kelly's Personal Construct Theory (PCT) (Kelly, 1955, 1969) which
suggests that people form idiosyncratic interpretations of reality based on a number
of dichotomous variables, referred to as personal constructs or attributes. A personal
construct is a bi-polar similarity-difference judgment. For example, when we meet a
new person we might form a construct friendly-distant to interpret her character. In
this process we perform two judgments: one of similarity and one of dissimilarity.
Both judgments are done in comparison to reference points: people that we regard
as friendly or distant.
To elicit the idiosyncratic attributes of each individual, the RGT employs a tech-
nique called triading , where the participant is presented with three products and is
asked to “think of a property or quality that makes two of the products alike and
discriminates them from the third” (Fransella et al., 2003). Once a bipolar con-
struct is elicited, the researcher may further probe the participant to elaborate on
the construct through the laddering and the pyramiding techniques (see Fransella
et al., 2003). Laddering seeks to understand what motivates a given statement and
thus ladders up in an assumed means-ends-chain (Gutman, 1982) towards more ab-
stract qualities of the stimuli; in laddering the researcher typically asks the partici-
pant whether the mentioned quality is positive or negative, and subsequently probes