I ran an analysis on SPSS talking about the mixed attitude( positive -Negative-neutral ) among males' and females' who experienced to drinks (beer-wine-water)after watching an advertisement about the differences of these three kinds of drinks and their imagery about that difference

Levene's Test of Equality of Error Variances Levene Statistic df1 df2 Sig. Beer + Corpse Based on Mean 1.305 18 268 Based on Median 1.259 18 .277 Based on Median and 1.259 15.367 .279 with adjusted df Based on trimmed mean 1.328 18 264 Beer + Sexy Based on Mean 1.009 18 328 Based on Median 865 18 .365 Based on Median and .865 18.000 365 with adjusted df Based on trimmed mean 982 18 .335 Beer + Person in Based on Mean 1.813 18 .195 Armchair Based on Median 1.924 18 .182 Based on Median and 1.924 17.204 .183 with adjusted df Based on trimmed mean 1.842 18 191 Wine + Corpse Based on Mean 1.048 18 320 Based on Median 621 18 .441 Based on Median and .621 17.945 441 with adjusted df Based on trimmed mean 1.001 18 330 Wine + Sexy Based on Mean 2.017 18 173 Based on Median 1.963 18 .178 Based on Median and 1.963 12.130 .186 with adjusted df Based on trimmed mean 2.012 18 173 Wine + Person in Based on Mean 071 18 793 Armchair Based on Median 005 18 .943 Based on Median and 005 15.198 943 with adjusted df Based on trimmed mean 060 18 810 Water + Corpse Based on Mean 804 18 382 Based on Median 763 18 394 Based on Median and .763 16.131 .395 with adjusted df Based on trimmed mean 824 18 .376 Water + Sexy Based on Mean .317 18 580 Based on Median 082 18 778 Based on Median and .082 17.998 778 with adjusted df Based on trimmed mean .310 18 585 Water + Person in Based on Mean 1.813 18 195 Armchair Based on Median .718 18 408 Based on Median and .718 13.534 .412 with adjusted df Based on trimmed mean 1.595 1 18 .223 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + sex Within Subjects Design: Drink + Imagery + Drink * Imagery