Were there a well accepted cognitively and neurologically plausible theory of how attention to affectively valenced information is affected in depression, understanding the above results could be a simple process of reconciling differences in experimental methodologies such that they could be explained by this theory. In reality a multiplicity of theories of attentional bias in depression exist, many of which predict different results on tasks such as the lexical decision task. Thus, it is difficult to determine which theories are supported or disconfirmed by the data presented above. Rather a careful analysis of these theories as suggesting multiple variables which could affect performance on the task may be used to explain the wide variation in observed results. From this analysis, an integrative theory accounting for the varying results in the meta-analytically aggregated studies will be proposed which suggests that delays on the lexical decision task in people who are depressed are a product of a number of processes acting to deflect cognitive resources away from the identification of a particular negative stimulus, while facilitation towards those same words may occur in people for whom the stimuli are depressotypic. To fully understand this theory it is useful to develop the history of cognitive models leading to the present formulation.