The Efficacy of Attentional Bias Modification for Anxiety: A Registered Replication

Generalised anxiety disorder is a prevalent condition linked to the presence of cognitive biases, including attention bias. Attention bias is the tendency to attend preferentially to threat-related stimuli and has been consistently observed in high anxious samples. Interventions to modify these biases have been developed with the aim of alleviating anxiety symptoms. However, while initial studies were promising, the current evidence-base for attention bias modification (ABM) procedures to alleviate symptoms is mixed. Furthermore, concerns have been raised regarding the potential for demand effects to be underlying previous significant findings. We revisited the efficacy of ABM, conducting a replication of a seminal study showing successful alleviation of anxiety symptoms following multi-session ABM training (Hazen et al., 2009). As a secondary aim, we quantified the potential influence of demand effects in the paradigm. Participants (N = 104) classed as high worriers attended seven lab-based sessions. This included five probe-based ABM training sessions (or sham-training control), and a pre- and post-training session in which levels of attention bias, worry, trait anxiety and depression were assessed. We adopted a Bayesian approach to analyses with a series of Bayesian t-tests evaluating the change in each outcome measure from pre- to post-training. These analyses revealed sensitive evidence for the null hypothesis of no effect for all measures except the measure of depression (which was insensitive), and no sensitive evidence for demand effects. These findings join a growing body of literature suggesting that probe-based ABM training does not lead to a robust reduction in bias or anxiety. Therefore, alternative ABM paradigms should be investigated.