Are there two independent evaluative conditioning effects in relational paradigms? Dissociating the effects of CS-US pairings and their meaning

ABSTRACT

Recent research into evaluative conditioning (EC) shows that information about the relationship between the conditioned and unconditioned stimuli can exert strong effects on the size and direction of the EC effect. Additionally, the co-occurrence of these stimuli seems to exert an orthogonal effect on evaluations. This finding has been interpreted as support for two independent types of EC effects. However, previous research devoted to this question relied on aggregated evaluative measures, allowing for alternative interpretations. In four experiments, we developed and validated a multinomial processing tree model that distinguishes effects of the pairings from effects of the meaning of the pairings. Our findings suggest that two independent EC effects contribute to overall evaluative change in a relational EC paradigm. The model that we developed offers a helpful method for future research in that it allows for an assessment of the effects of manipulations on processes rather than overall performance on an evaluative measure.

KEYWORDS:

• Evaluative conditioning
• attitudes
• dual-process models
• process dissociation
• multinomial processing tree models

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Note that our decision to estimate the p-parameter conditional on the absence of memory for the meaning is merely intended to reflect the structure of the memory task. We do not make any claims about the hierarchical nature of the processes involved in this decision: An MPT model that estimates the m-parameter conditional on the absence of a response in line with the pairing results in similar branch probabilities, even though the order of parameters is different.

2 Participants also completed an affective misattribution paradigm (AMP; Payne, Cheng, Govorun, & Stewart, 2005) in Experiments 1 and 2. In Experiment 1, the AMP was completed right before the memory task. The AMP was the last task in Experiment 2. In both experiments, we did not obtain effects of the pairing or relations on responding. When combining data from both studies, we observed a small and marginally significant effect of the pairing on choices: Chinese ideographs that were preceded by positively paired CSs were more often classified as “positive” compared to Chinese ideographs that were preceded by negatively paired CSs (4 percent point difference). Note, however, that the task order differed between experiments. Therefore, it is unclear whether merging the data is justified.

3 The predictive value of the MPT parameters for evaluative change also did not depend on this decision.

4 Using the latent-trait approach, we assumed that (transformed) individual parameter estimates follow a multivariate normal distribution. Individual parameter estimates showed indications of bimodality, potentially violating this assumption. For this reason, all MPT models were also estimated for each participant separately (without assuming hyper-distributions) and with the latent-class approach (Klauer, 2006). For all experiments, parameter estimates did not substantially deviate between the three methods. See Supplement B for a detailed explanation of the conducted analyses.

5 Evaluative change was predicted with a mixed model similar to the mixed models described in Experiments 2–3, with the exception that “instruction” was included as a between-participant factor. However, Experiment 4 was powered on detecting a difference between the instruction conditions using MPT models. In result, the sample size was relatively small to investigate interactions between within- and between-participant factors using a mixed model. In addition, most analyses conducted in Experiment 3 were not the primary focus of Experiment 4. For these reasons, we abstain from describing these analyses in detail. Instead, a complete description of the analyses conducted in Experiment 4 can be found in Supplement C. The pattern of results obtained in Experiments 2–3 also held in Experiment 4.

6 See Supplement C materials for the results of the hierarchical MPT approach.

Additional information

Funding

The research reported in this article was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – GRK 2277 “Statistical Modeling in Psychology” and projects HU 1978/ 4-1 and HU 1978/ 7-1 awarded to Mandy Hütter.