Results indicate that depressed research participants were delayed on positive words in the valence identification task, and slightly delayed on negative words in the lexical decision task with respect to nondepressed participants. These results are robust under a number of methods for aggregating reaction times. Additionally reaction times do appear to be somewhat linearly biased as a function of depressive symptomatology, though systematic biases are most apparent when individuals are broken into ``depressed'' and ``nondepressed'' groups as a function of having an arbitrarily high number of symptoms associated with clinical depression. Still, the largest portion of the variation in lexical decision times was accounted for not by depression, but by biased performance on the valence identification task. The small magnitude of the interaction of valence identification with depression suggests that a ruminative coping style, independent of depression is a primary factor in biasing information processing. Similarly, the lack of interactions with previous depression argue against a depression-mediated characterological formalism for rumination.
Results are thus somewhat consistent with the affective interference theory. The findings suggest that delays found on the lexical decision task are the result of interference by cognitive processes used to attend to affective valence. Support is provided for the simultaneous, interactive determination of affective and lexical content. Some instances of depression may be understood as involving overattention to the affective valence of negative stimuli within the cognitive-affective ruminative loop, while others, presumably, evidence themselves in other manners.