The principal limitation of this study was that many of the measures employed to predict both competence-related behaviour and expectations for competence, were concomitantly being assessed as to their ability to measure forms of stereotyping. As such no conclusions could be drawn as to the nature of processing involved in generating expectations for competence.
According to status characteristics theory, as it is the case that endorsing the superior state of the characteristic (i.e., man), is reflected by endorsing the inferior state (i.e., female), then beliefs about the one state should correlate with beliefs about the other. This was supported by the fact that no difference was obtained between male and female items in both the WrdTask and the SentComp (the two proposed measures of automatically activated implicit knowledge), and as such it can be assumed that these stereotypes were equally available. However, in both the OthsVws and the YrVws the male and female subscales were not highly correlated. Thus as the focus of the behavioural task was upon men's perceptions of women's competency, the relationship between stereotyping and competence-related behaviour may have been more appropriately tested if the target in each of the stereotyping measures reflected the target in the behavioural task.