Approach action tendencies toward positive stimuli and avoidance tendencies from negative stimuli are widely seen to foster survival. Many studies have shown that approach and avoidance arm movements are facilitated by positive and negative affect, respectively. There is considerable debate whether positively and negatively valenced stimuli prime approach and avoidance movements directly (i.e., immediate, unintentional, implicit, automatic, and stimulus-based), or indirectly (i.e., after conscious or non-conscious interpretation of the situation). The direction and size of these effects were often found to depend on the instructions referring to the stimulus object or the self, and on explicit vs. implicit stimulus evaluation. We present a meta-analysis of 29 studies included for their use of strongly positive and negative stimuli, with 81 effect sizes derived solely from the means and standard deviations (combined N = 1538), to examine the automaticity of the link between affective information processing and approach and avoidance, and to test whether it depends on instruction, type of approach-avoidance task, and stimulus type. Results show a significant small to medium-sized effect after correction for publication bias. The strongest arguments for an indirect link between affect and approach-avoidance were the absence of evidence for an effect with implicit evaluation, and the opposite directions of the effect with self and object-related interpretations. The link appears to be influenced by conscious or non-conscious intentions to deal with affective stimuli.

Attractive cues have been shown to evoke automatic approach biases in tasks such as the Automatic Approach Task or Stimulus Response Compatibility task. An important but as yet not studied question is the role of temporal dynamics in such tasks: the impact of automatic processes may depend on the interval between cue and response. The current proof of principle study tested this hypothesized time-dependence of the approach bias. Secondary goals included the exploration of effects of alcohol cues and virtual hand stimuli. 22 participants performed an SRC task in which the delay between the presentation of the cue and the possibility to select the response was manipulated. Results revealed an approach bias that decayed over longer delays. Thus, the approach bias was indeed dependent on processes that are transiently evoked by cues. The results did not show significant effects of alcohol cues or a virtual hand. Temporal dynamics may be an essential feature of approach biases.