Image credit: EASP
Abstract
Approach/avoidance responses can be crucial for survival. The literature reports approach/avoidance compatibility effects: participants are faster to approach positive stimuli and to avoid negative ones than the reverse (e.g., Solarz, 1960). Relying on a grounded cognition approach, authors reasoned that the visual information of whole-body approach/avoidance movements should be crucial in producing approach/avoidance compatibility effects (Rougier et al., 2018). This grounded cognition approach suggests that the sensorimotor information of our past experiences is encoded within memory traces (Versace et al., 2014). Thus, memory traces corresponding to past approach/avoidance experiences should include the visual information of approach/avoidance, that is an increase/reduction of stimuli sizes on the retina as well as modifications of the whole visual environment due to modifications of the observer point of view in the environment. Accordingly, experiments implementing this kind of visual flow allowed to observe large and replicable compatibility effects. Nevertheless, these experiments relied on the inference that such a visual flow of approach/avoidance was actually associated in memory traces with positive/negative words respectively.
The aim of our studies was to get rid of this inference by manipulating experimentally approach/avoidance experiences. In a first experiment, 158 participants repeatedly approached or avoided two groups of novel stimuli. Critically, the approach/avoidance movements were simulated through a whole-body visual flow of approach/avoidance. In the second phase of this experiment, participants were primed with a previously approached/avoided stimulus for 200ms and had to approach/avoid a square or a diamond displayed 100ms later. We predicted and observed a significant compatibility effect: when primed with a previously approach/avoided stimulus, participants were faster to respectively approach/avoid geometric shapes compared to the opposite, t(154) = 5.22, p = .025, dz = 0.42, IC 95% [0.25; 0.58]. Three outliers were excluded, and the reported p-value was corrected for sequential testing following Lakens (2014)’s recommendations.
One limitation of this study was that the instructions in the first phase concerned directly the group of the stimuli. Thus, we conducted another study in which instructions in the first phase concerned the stimuli color rather than their group membership. We predicted and observed, based on the data of 79 participants, a significant compatibility effect, t(76) = 2.676, p = .009, dz = 0.30, IC 95% [0.08; 0.53]. Note that we excluded two outliers in this analysis. These two studies indicate that we are faster to approach and to avoid after seeing a previously approached and avoided stimulus.
Yoann Julliard
Post-Doc
My research interests include ‘implicit’ processes, grounded cognition, methods, and data analysis.