Ambivalent sexism: research into attitudes towards women

Many of us are familiar with the concept of misogyny: hatred of women. Sexism has another face, though: the belief that women are wonderful and must be protected from the big, bad world.

These two sides to sexism were given a name in a paper by Glick and Fiske (1996): ambivalent sexism. Ambivalent sexism consists of two types of attitude towards women: hostile sexism, and benevolent sexism. Hostile sexism is classic prejudice; benevolent sexism is the view that women are lovely, fluffy nurturing caregivers (or, as the paper puts it, intimacy-seeking and prosocial). Within these categories are three “sources” of ambivalent sexism, each with its corresponding hostile and benevolent face.

First, paternalism. Paternalism is theorised to come in two forms. Dominant paternalism is the idea that men should control women, while protective paternalism is the notion that men should protect women.

Second, gender differentiation. Competitive gender differentiation is a set of beliefs that bolster the idea that men are the better sex, while complementary gender differentiation, its benevolent counterpart, focuses on the “equal but different” myth, wherein women have their own, special roles in the kitchen.

Finally, heterosexuality. The theory of ambivalent sexism acknowledges that a major source of sexism is the hegemonic heterosexual ideal. Heterosexual hostility is the viewing of women as sex objects and fear of female sexual power, while intimate heterosexuality romaticises this objectification and sees men as incomplete without a woman.

The theory therefore provides a fairly comprehensive account of sexism. It does not just stop at theorising.

Measurement of ambivalent sexism

Ambivalent sexism is measured by a questionnaire called the Ambivalent Sexism Inventory (ASI). This measure was subjected to the rigorous development and validation standards typically used in questionnaire development (for those interested in methodology, it is described fully in the Glick and Fiske paper, which is available in full without paywall).

The ASI consists of 22 items; 11 in each category. Examples of questions which tap hostile sexism are “Once a man commits, she puts him on a tight leash” or a reverse-scored item “Feminists are not seeking more power than men“. Reverse scoring allows researchers to check if participants are just selecting the same response for every item on a questionnaire, and also help to test the reliability of the measure.

Examples of questions which tap into benevolent sexism include “A good woman should be set on a pedestal“, “Women have a quality of purity few men possess“, and “Men are complete without women“. Spot the reverse-scored item.

The Glick and Fiske study found that benevolent and hostile sexism were distinct, but they were also correlated with one another, suggesting that people who hold hostile sexist attitudes also hold benevolent sexist attitudes.

A problem with the ASI, though, is that it is dependent on self-reporting. Even in an anonymous questionnaire, research participants may give responses that make them seem socially desirable (i.e. less of a sexist knobend). Furthermore, a questionnaire may influence their behaviour or responses to other questions if the participant guesses that the study is about sexism. For that reason, some researchers prefer to modify the ASI to present scenarios or observe behaviour.

Effects of ambivalent sexism

Much of the ambivalent sexism research has focused on workplace sexism. Hostile sexism has been linked to negative evaluations of women candidates for a managerial job and higher recommendations for a male candidate for the same role. It has also been linked to greater tolerance of sexist events after hearing a sexist joke, which suggests that sexist humour does have real-world implications, for hostile sexist people, at least.

Benevolent sexism has many real-world implications. It, too, has been linked to low evaluations of women in the workplace, as women are seen to be neglecting their traditional roles as caregivers and homemakers. As well as this systemic negative effect on women, the impact of benevolent sexism extend to psychological effects. When experiencing benevolent sexism, women perform worse at various cognitive tasks, which suggests that the benevolent sexist attitude further reinforces a vicious circle which allows women to do worse.

What about teh menz?

A similar scale has been created for measuring attitudes towards men: the Ambivalence Towards Men Inventory, a 20-item questionnaire which also differentiates between hostile and benevolent attitudes. This measure has generated much less research than the ASI, with fewer real-world implications. However, using the measure, it has been found that feminists are not man-haters: in fact, women who identify as feminist score significantly lower on hostility towards men.


While the problems of hostile sexism are well-known, and generally viewed as less acceptable in our society, “benevolent” sexism, too, has huge implications for equality for women. Benevolent sexism still allows women to be viewed as objects, and unworthy of equal employment, yet it is thoroughly acceptable to express opinions that women are cute little walking wombs. This needs to change.

32 thoughts on “Ambivalent sexism: research into attitudes towards women”

  1. “The Glick and Fiske study found that benevolent and hostile sexism were distinct, but they were also correlated with one another, suggesting that people who hold hostile sexist attitudes also hold benevolent sexist attitudes.”

    Possibly as a Madonna/Whore complex?

    1. Absolutely: one of the studies Glick and Fiske did suggests that men who score high on both hostile and benevolent sexism tend to sort women into “homemakers” and “career women”–I haven’t seen much on sexual attitudes, but like you, I’d guess that this categorisation happened there, too.

  2. Nice to have a citation on feminists officially not being manhaters!

    Now to copy/paste it all over every newspaper discussion board and the ridiculous straw-feminism so many commentors seem convinced is taking over teh worldz…

  3. here’s the thing. You know who I think are the most ‘benevolently sexist’? Feminists.

    I suggest an experiment. Go to Cath Elliott’s last blogpost. And read the last few comments about the cif online discussions. What the women on Cath’s Blog are saying is that women who comment on cif need special protection by the ‘cif moderators’ from nasty, hostile sexist men. That is asking for paternal protection as you describe it in this post.

    I also often read feminists actually saying that women are superior too and not as nasty/violent/dangerous/powerful than men, as people. This suggests feminists have a ‘sexist’ but ‘benevolent’ attitude towards women.

    Then take laws like e.g. anonymity for accusers in rape cases. This suggests those women need ‘special’ protected treatment from the ‘paternalistic’ state.

    Then come back and we can discuss if feminists hate men too.


    1. This isn’t the case: there have been a number of studies using the ASI on women, and none have found relationships with feminist attitudes–in fact, benevolent sexism is associated with victim blaming in rape cases and handwaving away inequality. It’s fallacious, therefore, to say that it’s associated with feminism. Perhaps you have misunderstood the literature or the nature of feminism 🙂

      1. why do feminists always use the term ‘fallacious’ is it because it sounds like fellatio?

        I think I have read the study and feminism quite adequately thanks. You have not come up with any argument against my points. You have just mentioned the findings of the study and listed other examples of ‘benevolent sexism’.

        1. Your point is, essentially a hypothesis: that feminists display higher levels of benevolent sexism than non-feminists.

          There are several ways of testing this hypothesis: firstly, by administering the ASI to a group of feminists and a group of non-feminists and see if they differ in levels of benevolent and hostile sexism in a cross-sectional study. For a slightly more rigorous experimental design, one may wish to conduct a longitudinal study, perhaps by exposing a group of participants to feminist ideas, measuring benevolent sexism before and after this exposure. To make things even more rigorous, the researcher might want to also include a control group of participants who have not been exposed to feminist ideas. Per your hypothesis, one would expect to see an increase in benevolent sexism following exposure to feminist ideas.

          This study was conducted and the opposite emerged: benevolent sexism decreased, as did hostile sexism.

    1. That post, unfortunately, rather misunderstands the point of trigger warnings. Trigger warnings are for people who experience PTSD following a traumatic event (e.g. a rape, violence, bullying), so that they can avoid a triggering of their condition. It is perfectly in keeping with the Social Model of Disability, although some may believe that to be too politically correct.

    2. Old post, but need to stick my oar in here; stavvers is totally correct. Trigger warnings are just as much for men who have experienced traumatic events as for women. They have nothing to do with gender and everything to do with consideration for other human beings. I spend a lot of time hanging round online support blogs for sexual abuse survivors, and trigger warnings are plastered all over them, with good reason.

    1. thanks Emma I didn’t realise this was a wrestling match. Isn’t that a bit of a violent metaphor for a conversation?

      I don’t do psychological studies to prove every hypothesis I make about life and neither do you Zoe. I reject this kind of epistemology in terms of measuring how ‘sexism’ works as well. As I don’t think questionnaires like this actually measure power dynamics in society accurately.

      You have just done a post about Cameron’s ‘casual sexism’ in the House of Commons. Feminists are rushing to state their ‘hypothesis’ that this demonstrates the sexism of the Tories and the sexist culture of the House of Commons. Nobody is filling in a questionnaire to see how everyone feels about this before you all reach your conclusions.

      I know you are a psychologist and a positivistic one at that. I am not into that way of examining human relations.

      P.s. I know what trigger warnings are and i think they treat women like children.

      1. Your thoughts about benevolent sexism–which was clearly marked as a psychological theory–were unfounded and in conflict with research into the field. This is therefore thoroughly unrelated to the Cameron post, which was not clearly marked as an overview of a psychological theory.

        The ASI is a measurement of attitudes, not power relations. It therefore does take an individual rather than systemic perspective on sexism, which is a valid epistemological point. There are other branches which look at prejudice from the systemic perspective, which I will no doubt write about at some time, but for the present, I think it’s fair to say that feminists do not display higher levels of benevolent sexism.

        Tsk. Surely trigger warnings treat everyone like children rather than just women, from that perspective?

      2. Thing is, QRG, trigger warnings aren’t just aimed at women. As a man who suffers from severe depression, I appreciate trigger warnings for things like suicide or self harm so that if I’m in an already vulnerable emotional state I don’t come into contact with material which might be detrimental to my health. I don’t see that as being ‘treated like a child’, I see it as something that allows me to make informed decisions related to my health. The same applies to ‘feminist’ trigger warnings. They’re not a feminist thing, they’re a mental health thing.

  4. I couldn’t resist…

    Re: Trigger warnings.

    Most people who are afflicted with depression, being obese, being anorexic, having been raped, being bullied, etc. generally have to navigate life somehow. If being triggered by a website which will USUALLY discuss topics of that nature sets you off, then you are just basically non-functional and shouldn’t even be on the internet. In fact, you should just act under the presumption that potential triggers are everywhere!

    I don’t know why it’s anyone’s obligation to coddle and accommodate for someone’s hypersensitivity. Unfortunately, no one thinks to themselves as they go about their everyday tasks and engage in everyday interactions that they have to accommodate for potential victims. By the way, I am slapping one big trigger warning on my whole blog, so I wouldn’t advise more sensitive members to visit.

    This post portrays trigger warnings and the levels of ridiculousness it can escalate to quite well:

  5. It is mainly feminist blogs and spaces who use trigger warnings. I have mainly seen women refer to them and call for them. I have been accused of ‘triggering’ women just by making a pretty basic reasoned comment.

    ‘The ASI is a measurement of attitudes, not power relations’

    well I am a Foucauldian and I think everything operates within power relations. I thought even feminists thought gender was always about power.

  6. I have to say Los, I agree with Sofia. I have had depression in my life, and ‘post traumatic stress’ as a result of a domestic violence situation, but I didn’t expect other people to change how they communicated or behaved towards me as a result.

    I find the News one of the most upsetting things to watch. I don’t expect them to put a trigger warning on the news! I just don’t watch it if I am feeling sensitive.

    I think we have to learn to manage our own issues.

    The thing about feminists and ‘trigger’ warnings is they seem to think feminist women are never responsible for their own (mental)well-being. It is always some other person(usually a man)’s fault if they are suffering.

    I think this is relevant to those psychology tests because the ‘attitudes’ they have included are pretty cliched. There are other attitudes eg held by many feminists that may indicate a ‘benevolent sexism’ but they are not included in the questionnaire.

    1. It’s entirely possible these other attitudes were included in the original 140, but distilled because they mapped highly in similarity to final 22 questionnaire items.

      Nobody wants to fill in a 140-item questionnaire. Nobody wants to analyse a 140-item questionnaire when a 22-item one does the job just as well!

      1. You have no evidence of that stavvers! I thought you were all about the evidence.

        I doubt ‘feminist attitudes’ of paternalism and protection were included as these rarely get criticised. Hence when people like me criticise them we get all manner of shite thrown at us.

        1. Knowing how questionnaire data reduction is done, it’s a fairly valid guess. 😉

          Protective paternalism is one of the domains on the questionnaire, so perhaps I’m misunderstanding, but I imagine a lot of the items you believe to be missing are covered by the item ‘Women are too easily offended’ (which loads on to HS in the factor analysis).

  7. No I don’t really mean being easily offended. I mean thinking they need special treatment and protection.

    I think the very fact ‘trigger warnings’ exist shows women feel in need of being protected from the big bad world of men. And, like I said upthread, that latest discussion on Cath Elliott’s blog is FULL of women calling for special treatment, e.g. on cif discussion boards. One idea was to make discussions of feminist articles on cif ‘women only’!

    1. Two issues here:
      1. That the main rationale for trigger warnings comes from a mental health perspective, not to protect women’s feelings.
      2. That Cath Elliott is not representative of all feminists. As someone who has read up on feminist, I’m sure you’re aware we’re hardly a homogenous mass!

      1. But ‘mental health’ relates to ‘feelings’ madame psychologist.

        I think feminists all have some things in common. And one thing they have in common is how they respond to robust critique of their ‘ideas’. But feminism hasn’t had any ideas as far as I can see since the 1980s, Cordelia Fine excepted.

  8. Also, a woman seeing a man as a sex object even faces hostile sexism because she is not behaving according to her role.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.