Emotional arousal does not modulate stimulus-response binding and retrieval effects

ABSTRACT

The adaptation-by-binding account and the arousal-biased competition model suggest that emotional arousal increases binding effects for transient links between stimuli and responses. Two highly-powered, pre-registered experiments tested whether transient stimulus-response bindings are stronger for high versus low arousing stimuli. Emotional words were presented in a sequential prime-probe design in which stimulus relation, response relation, and stimulus arousal were orthogonally manipulated. In Experiment 1 (N = 101), words with high and low arousal levels were presented individually in prime and probe displays. In Experiment 2 (N = 170), a high arousing affective word was presented simultaneously with a neutral word during the prime display; in the subsequent probe display, either the arousing or the neutral word repeated or a different high versus low arousal word was shown. Data from both experiments did not demonstrate a modulation of SRBR effects by stimulus arousal and SRBR effects were of equal magnitude for word stimuli of high and low arousal levels. These null results are not in line with binding accounts that hypothesise a modulatory influence of emotional arousal on perception-action binding.

KEYWORDS:

• Stimulus-response binding
• event files
• stimulus arousal
• episodic retrieval
• action control

Acknowledgements

We thank Georg Reinhardt for help with programming the experiments. This research was supported by grants from the German Research Foundation (GI 1295/2-1) to Carina G. Giesen.

Author note

Carina C. Giesen, Friedrich-Schiller-Universität Jena, and Andreas B. Eder, Julius-Maximilians-University Würzburg, Germany.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The LMM analysis was only preregistered for Experiment 2. However, we decided to use the same data-analytic approach for both studies given the methodological strengths of LMM and for a better comparability of results across experiments. Preregistered analyses of Experiment 1 using an ANOVA approach are reported in the Supplement.

2 A 2 (arousal level) × 2 (word relation: identical repetition vs. word change/same arousal category) × 2 (response relation) ANOVA yielded a significant main effect of Response Relation, F(1, 100)    251.59 p < .001, η_p² = .72, and a significant Word Relation × Response Relation interaction effect, F(1, 100)  = 77.94, p < .001, η_p^2 = .44. Follow-up analyses on the Word Relation x Response Relation interaction revealed a disordinal interaction effect pattern typical for the retrieval of stimulus-response bindings: Word repetition (compared with word changes) produced a significant performance benefit for response repetition probes of about Δ = 18 ms, t(100) = 9.56, p<.001, d_z= 0.96. However, for response change probes, word repetition (compared with word change) produced a significant performance cost of about Δ =  -7 ms, t(100) = 4.04, p<.001, d_z= 0.40 (see also the ANOVA results of Experiment 1 in the supplement).

3 A 2 (arousal level) × 2 (word relation) × 2 (response relation) ANOVA yielded a significant main effect of Response Relation, F(1, 169)    325.04, p < .001, η_p² = .66, and a significant Word Relation × Response Relation interaction effect, F(1, 169)    52.64, p < .001, η_p² = .24. Follow-up tests indicated that word repetition (compared with word changes) produced a significant performance benefit for response repetition probes of about Δ = 7 ms, t(169)    5.85, p < .001, d_z = 0.45. However, for response change probes, word repetition (compared with word change) produced a significant performance cost of about Δ =  -6 ms, t(169)    5.07, p < .001, d_z = 0.39.

Additional information

Funding

This work was supported by Deutsche Forschungsgemeinschaft: [Grant Number GI 1295/2-1].