The word-rating questionnaire, administered in the last semester of the experiment, should provide explicit indications of the intensity of negativity or positivity of words for individuals. The affective interference theory predicts that ratings of high negativity for words will be associated with facilitation in the lexical decision task, while ratings of lower negativity will be associated with relative delays on that task. Thus, correlations between rated negativity and reaction time on the lexical decision task may be determined for each person. The magnitude of the average of the z-transformed correlations will allow a determination of whether people are, in general, more delayed for more depressotypic cognitions. a regression using these scores as dependent variables, and depression as an independent variable will allow determination of whether such a finding varies with depression.
Correlations were performed with BDI score and z-transformed
correlations of lexical decision times with word ratings. Only for
positive words in the 100 ms condition did depression account for over
4
of the variation in the lexical-decision-word-rating
correlations. Similarly a 3 (valence: Positive, Negative, Neutral) x 3
(SD: 50ms, 100ms, 150ms) x 2 (Depression: Depressed, Nondepressed)
split plot multivariate ANOVA revealed no statistically significant
main effects or interactions with the depression variable.