After noting that the competence-related behaviour of influence rejection was greater after exposure to the traditional subtype than after exposure to the modern subtype, it was speculated that according to the type of woman involved in the task-related setting, stereotypic processing may be activated differentially to affect competence-related behaviour. As the measures employed in this study predominantly contained material pertaining to the broader female stereotype (the traditional woman), and that features of this stereotype parallel those describing this study's traditional subtype, it was surmised that measures of stereotyping may predict behaviour in the traditional subtype condition.
In conducting a stepwise regression analysis on those within the traditional subtype condition (n = 20), with P(S) scores the dependent variable, and the nine stereotyping measures the independent variables, the correct response rate in the WrdTask emerged as a significant predictor F(1,18) = 5.20, p<0.05 (Table 14). In conducting a stepwise regression analysis with a low ratio of cases to measures, however, it was acknowledged that the results may have been somewhat spurious.
|Correct Response Rate (WrdTask)||0.76||0.47||0.47||0.22||5.20||0.03|
Note. The variables excluded from the model were OthsVwsM, OthsVwsF, stereotype consistent and inconsistent response latencies in the WrdTask, SentComp, YrVwsM, YrVwsF, ASIWSQ.
As is evident in Table 14, the correct response rate in the WrdTask explained 47% of the variance in the P(S) scores. Even though this solution may have resulted from the low ratio of cases to measures (Tabachnick & Fidell, 1996), it may suggest that after exposure to certain stereotype-related information, in this case information pertaining to a traditional female stereotype, the process entailed in generating expectations for competence occurs in a similar manner to the processes entailed in agreeing with the stereotype (correct response measure).