Kimberly A. Neuendorf, Thomas D. Gore, Amy Dalessandro, Patricie Janstova, and Sharon Snyder-Suhy have published Shaken and Stirred: A Content Analysis of Women’s Portrayals in James Bond Films at 62 Sex Roles 747-761 (2010). Here’s the abstract.
A quantitative content analysis of 20 James Bond
films assessed portrayals of 195 female characters. Key
findings include a trend of more sexual activity and greater
harm to females over time, but few significant across-time
differences in demographic characteristics of Bond women.
Sexual activity is predicted by race, attractiveness, size of
role, and aggressive behaviors. Being a target of weapons is
predicted by size of role, sexual activity, and weapon use,
while being harmed is predicted principally by role. End-offilm
mortality is predicted by sexual activity, ethical status
(good vs. bad), and attempting to kill Bond. This identification
of a link between sexuality and violent behavior is
noted as a contribution to the media and sex roles literatures.
The attempt to quantify attractiveness strikes me as particularly problematic (though I was surprised to see from Table 2 that apparently there was better agreement between coders on this than on hair length!). And it doesn’t seem terribly informative to say that size of role is a predictor of sexual activity; I wouldn’t expect those two variables to be independent in any case. Activity of any sort, including sexual activity, contributes to the size of the role.