As a result of the ANCOVAs, univariate tests showedthe main effects of ad type on the four dependentvariables: for Aad, F(2,320) = 21.078, p = .001; for Abrand,F(2,320) = 7.922, p = .001; for PI, F(2,320) = 6.694, p= .001; for MC, F(2,320) = 7.517, p = .001. The covariate,fanship, was significant for all four dependentmeasures: for Aad, F(1,320) = 6.646, p = .010; for Abrand,F(1,320) = 142.173, p = .001; for PI, F(1,320) = 83.268,p = .001; for MC, F(1,320) = 50.818, p = .001. Means ofthe four dependent variables, after controlling for theeffect of fanship, are shown in Table 3.