The “emotion misattribution” procedure: Processing beyond good and bad under masked and unmasked presentation conditions

Abstract

In general, it is assumed that misattribution in the Affect Misattribution Procedure (AMP) is restricted to crude affect due to its unbound nature, especially under limited presentation conditions. In two experiments, we investigated whether emotion-specific misattributions occur using a four-category misattribution procedure. Experiment 1 yielded emotion-specific misattribution effects under clearly visible presentation conditions demonstrating that the procedure is principally susceptible for emotion-specific effects. In Experiment 2, we employed masked presentation conditions impeding conscious prime perception. A specific pattern of emotion-specific misattributions effects emerged indicating some emotion-specific processing at initial stages of processing. However, not each emotion was misattributed equally. We discuss the implications of these results for the non-conscious processing of emotional information, for the supposed mechanisms of the AMP and its implicit nature.

Keywords:

• Emotion
• Misattribution
• Priming
• Masked presentation
• Subliminal

This research was supported by a grant from German Research Foundation [DFG-WE 2284/9-1] and the Netherlands Organization for Scientific Research [NWO; DN 57-239] to Dirk Wentura and Juliane Degner.

This research was supported by a grant from German Research Foundation [DFG-WE 2284/9-1] and the Netherlands Organization for Scientific Research [NWO; DN 57-239] to Dirk Wentura and Juliane Degner.

Notes

¹ Analyses with the neutral prime condition as baseline condition yielded essentially the same results. Given the interpretation bias for the neutral target faces, however, neutral primes might have also been interpreted as sad, thereby not providing an adequate neutral baseline condition.

² Data of seven further participants were excluded because the timing parameters of presentation accidentally differed from the preset ones.

³ There was one further difference between Experiment 2A and 2B: With Experiment 2B, we made an (unsuccessful and post hoc not very plausible) attempt to subtly modify subjective control expectations using a judgement of contingency paradigm (JOC; see Alloy & Abramson, 1979; Kaufmann, 2009). Before the misattribution procedure, participants were presented with 40 trials of a JOC task. In each trial, participants were instructed to decide whether they want to click a button or not. After that, either a blue circle appeared on the screen or not. Participants' task was to figure out to what degree they had control over the appearance of the circle. Actually, participants could not influence the appearance by their behaviour. Only the frequency of appearance was varied between participants (75% of trials versus 25%). This subtle manipulation has been shown to influence the perceived control over the appearance of the circle (Kaufmann, 2009). Indeed, participants in the 75% condition reported significantly higher subjective control over the appearance of the circle compared to participants in the 25% condition. Nevertheless, the manipulation did not have any impact on the priming effects in the misattribution task (Fs < 1 for the moderations of the constant effect and the emotion effect). This result and the fact that Experiment 2B yielded almost the same results as Experiment 2A lead to the conclusion that there was no transfer of the JOC task to the misattribution task.

⁴ These five participants all had outlying values on the prime detection index according to Tukey (1977; i.e., values were more than 1.5 interquartile ranges above the 75th percentile).