Attention bias modification (ABM), in which participants are trained to direct attention away from negative information, has been shown to reduce anxiety. However, such findings have been inconsistent. Changes in attentional bias are often absent, suggesting need for further investigation of the underlying mechanisms of ABM, as well as better statistical methods to analyze ABM data in order to reduce inferential error. In this study, we (a) compared inhibition control training to standard ABM training conditions, and (b) demonstrated the benefits of using simple Bayesian analyses to analyze ABM data. We recruited 116 participants and assessed their attentional bias prior to and after training, which involved practice avoiding negative stimuli, attending to negative stimuli, or avoiding a non-emotional, exogenous attentional cue (inhibitory control training). Our results suggested no impact of any of the training conditions on attentional bias. We further demonstrate Bayesian analyses may help control for both Type I and Type II error relative to a frequentist approach.