Training Less Threatening Interpretations Over the Internet: Impact of Priming Anxious Imagery and Using a Neutral Control Condition

Cognitive Bias Modification to reduce threat interpretations (CBM-I) is a computer-based paradigm designed to train a less negative interpretation bias that has shown some success in the lab, but results for web-based CBM-I are often mixed. To test possible explanations for the poorer results online, participants high in social anxiety (N = 379) were recruited from Amazon’s Mechanical Turk to complete a single-session, proof-of-principle study to investigate: (1) whether web-based CBM-I can shift interpretations of social situations to be less negative and reduce anticipatory social anxiety, (2) whether a common “control” condition used in CBM-I studies is in fact inert by incorporating an alternate control condition, and (2) whether priming anxious imagery prior to training moderates CBM-I’s effects. Participants were randomly assigned to one of three training conditions: all positive, half positive/half negative, or neutral unemotional scenarios. Participants also received an anxious or neutral imagery prime before training. Although results were somewhat mixed across outcome measures, findings generally suggested that participants exhibited less negative interpretations of ambiguous social scenarios following positive training with an anxious imagery prime. There was also some evidence that the neutral training condition was associated with less negative interpretations, and evidence that the half positive/half negative training condition led to the least anticipatory anxiety, especially when paired with anxious imagery. Findings are discussed in light of different training effects for near- and far-transfer outcomes.