Threatening Virtual Reality Environments Enhance General but not Specific Pavlovian-Instrumental Transfer

Abstract

Pavlovian-Instrumental Transfer (PIT) exemplifies how pavlovian motivational influences modulate goal-directed behavior yielding outcome-specific (specific PIT) and general (general PIT) transfer. General PIT is commonly interpreted as outcome-general invigoration and is sensitive to stress. However, human PIT research typically uses visual, appetitive procedures, whereas rodent PIT research often uses auditory cues, limiting translation. We tested whether cue modality along with additional factors such as immersive threat contexts modulate PIT, and whether virtual reality (VR) enhances general transfer. Across three experiments (N = 196), participants completed a PIT task: (1) 2D appetitive PIT with auditory vs visual cues (Experiment 1; n = 60); (2) VR PIT comparing appetitive (positive reinforcement) vs aversive (negative reinforcement; ‘zombie’) contexts (Experiment 2; n = 40); and (3) aversive VR PIT preceded by immersive compound threat scenarios (neutral, spiders, contamination) in individuals stratified by contamination fear (Experiment 3; n = 96). Specific and general PIT were computed from baseline-corrected response rates. Stress induction (Experiment 3) was assessed using photoplethysmography-derived heart-rate variability (HRV), salivary alpha-amylase (sAA), and self-report. Robust specific (ηp2 > 0.45) and general PIT (ηp2 > 0.68) were observed across all experiments. In Experiment 1, PIT magnitude did not differ by cue modality; specific PIT exceeded general PIT across auditory and visual conditions. In Experiment 3, threat scenario type and contamination fear did not significantly alter transfer effects. Nevertheless, increased stress indices were observed, including phase-dependent HRV changes, elevated sAA from pre- to post-test, and higher self-reported anxiety (with stronger subjective fear/disgust in the contamination condition). Across experiments, general PIT was larger in the VR studies than in the 2D study whereas specific PIT remained stable; however, VR was not manipulated independently of reinforcement context and other procedural differences, precluding strong causal inference about its effect on general PIT. Despite this, human PIT appeared robust across cue modality and reinforcement context and there was indication that VR immersion may selectively amplify general invigoration while sparing specific PIT across all three experiments, albeit with a small effect (pseudo-R² = .06).