Populations

• These effects will only be present for cognitively-mediated depressions, which are assumed to be a product of cognitive feedback between affective and semantic structures as described by Ingram (1984). More biological depressions such as those which are a product of bipolar disorder may not show the same results. As such, measures of depression such as the Beck Depression Inventory (BDI) may be highly correlated with the observed information processing biases, while other measures of depression may not be so highly correlated.
  
• These effects will be present in formerly depressed people if rumination on negativity is trait-like.
  
• As Nolen-Hoeksema (1987) has suggested that ruminative coping styles are mediated by gender, so too may depressive attention biases due to rumination be mediated by gender.
• A final prediction may be generated from the network theory regarding the performance of people who are not depressed on the tasks. Because feedback from the affective structures is hypothesized to aid in the determination only of stimuli with affective content, it may be that stimuli which are positive or negative will be identified faster than neutral stimuli by both nondepressed and depressed people, on both tasks. This prediction is similar to that made by Kitayama (1990) who suggests that ``preconscious'' processing of affective information guides semantic processing. Notably, this prediction is, indeed born out in the lexical decision data summarized in the meta-analysis above.
  
• The affective interference theory predicts that induced depression should have no effect on lexical decision because the induction has not created a maladaptive network of negative constructs. Such a finding is supported by Challis and Krane (1988) and Clark et al. (1983). It is not tested in the present experiments.