Patterns of mood change

Two independent sets of data were analyzed, each consisted of self-reports of mood using 10 scales devised by means described in the article. The two data sets were, respectively, based on: (1) unforced mood changes over a 30-day period; and (2) forced mood changes measured following administration of either elation- or depression-inducing statements. Correlation matrices, factor analysis and rearranged correlation matrices were presented to investigate unipolar, bipolar and circumplex hypotheses. The forced data set proved to be the simpler case, with two factors, and the apparent circumplex of the factor plot proved to be the case following rearrangement of the correlation matrices. The unforced data, though confirming the simpler solution in some respects, proved to be more complicated and more difficult to conform to a model, which is by definition two-dimensional. Therefore, natural covariation may be too complex for the circumplex model. However, there is still a possibility that a broad bipolar model can not only fit a circumplex but may have heuristic value