Would be interesting to see more detailed apparaisal about the
"Women are wonderful" effect mentioned by C.H. Sommers. It's quite interesting in that it could intuitively be subsumed by the
ambivalent sexism theory/measure, though whether this is the case isn't clear as it could actually be evidence against the theory.
The paper that seems to be basis for considering that there is a bias in favour of women without ambivlanece is
ARE WOMEN EVALUATED MORE FAVORABLY THAN MEN? An Analysis of A ffifudes, Beliefs, and Emofions.
In conclusion, our findings suggest that people evaluate women quite favorably as a general social category. Although our respondents evaluated men favorably as well, their evaluations of women were more positive than their evaluations of men when we consider their overall attitudes and the attributes that they ascribed to the sexes. This conclusion was not contradicted by any evidence for covert negativity toward women at the emotional level. Nor did it appear that the overall positivity toward women masked an unusual amount of ambivalence.
But I imagine the discrepency perhaps comes about how they measure ambivalence where the above paper reported using
Kaplan's technique. And I suppose a difference could be that the ASI necessarily measures people within it's conceptual framework whilst the other researchers relied on free-response measures in conjunction with a list of beliefs measure. Which would have the possible benefit of appraising people's personal beliefs more accurately. But then it could be that because it lacked theoretical background that it couldn't properly discern such things. The nature of such methods is something i'm unsure about since I see the intuitive appeal of allowing people to more freely choose things as one could be confined to a particular framing that wouldn't apply if it were broadened. Though it's also intuitive that by losing all boundaries in order to maintain free associations that one doesn't necessarily find much to further understanding. Because when we have preconceptions, even in useful measures, it does make sense we'll kind of find what we're looking for to some degree. But by stepping away from that framework one can find the results that are beyond the scope of that framing also.
Regardless, I think the ambivalent sexism scale touches on a compelling point that there isn't an isolated hostility but also a corresponding positive ideal.
A shame C.H. Sommers didn't go into that and instead made it about a factual claim of finding no sex differences on things like multitasking rather than directly considering the nature of attitudes towards men and women in some nuance. Perhaps even going as far as to criticize the ambivalent sexism inventory.
https://www.ethicalpolitics.org/ablunden/pdfs/For%20Ethical%20Politics.pdf#page90
-For Ethical Politics