Investigating the attitude towards ambiguity: Interindividual differences in automatic activations of evaluations of ambiguity

Abstract

Direct measures of (in)tolerance of ambiguity provided evidence for a variation in liking of ambiguity. Given the limitations of these measures, we developed a direct measure of attitude towards ambiguity. However, the main part of this thesis deals with the questions, whether there are interindividual differences in the automatic activations of evaluations of subjective ambiguity (single information triggers multiple distinct representations). We developed a database with norms for ambiguous and unambiguous German words. A subset of these words matched for several dimensions was used in two indirect measures. The Implicit Association Test (IAT) assessed the relative strength of associations of ambiguity and clearness with positive and negative valence, respectively. In the Evaluative Priming (EP) paradigm ambiguous and unambiguous primes preceded targets with positive or negative valence. This allows to draw inference about the automatic evaluations of the primes. In order to validate potential variances in the automatic evaluations of the primes (indicated by interindividual differences in latencies as a function of prime type and target type), information about the attitude towards ambiguity was used as a moderator. Apart from the first study, which investigated the relation of direct measures with the IAT and showed unrelatedness, all other studies used the EP paradigm, using either the IAT score (studies 2 – 4) or the induced associations of opposite valence with ambiguity and clearness (studies 5 – 6) as moderators. Studies 2 and 3 provided evidence for interindividual differences in the activation of evaluations of ambiguity via a three-way interaction of prime type, target type and IAT score. However, this three-way interaction was not found in the replication study 4. In study 5, there was an interaction of prime type, target type, and induction but opposite to the expected direction. The post-hoc explanation for the partial contrast effect was further investigated by manipulating the SOA in study 6. However, in this study the induction had no influence on latencies in the EP paradigm, but there was a prime type, target type and SOA interaction. This can be explained in terms of contrast (long SOA) and assimilation (short SOA) effects if we consider the evidence across the aforementioned studies showing that, on average, participants had a stronger association of ambiguity with negative valence and clearness with positive valence. Summarized, evidence for interindividual different automatic evaluations of ambiguity was weak, but the results of the EP paradigms indicated a more negative (or less positive) automatic evaluation of ambiguity compared to clearness. The implications of automatic evaluations of the mental representation of ambiguity are discussed.