Positive Outcomes of Sexual Harassment?

04 September 2019


only yes is yes


In my previous article, Online Interactions and Harassment (23/04/2019) I explained the study that I was conducting and its related survey. Now the results are in! So what do they tell us?

Well, lots actually, but, as with so much research, they open up a whole new range of questions to look into. Firstly, let me sum up what this is all about: Sexual harassment, whilst detrimental psychologically, economically, socially, and sometimes physically, has a plethora of grey areas. Different people accept different behaviours in different situations. So how does one know, where, when, who…?

Perhaps the obvious answer would be to avoid all behaviour that may be interpreted as sexual. Bye-bye sex life and, eventually, mankind! Maybe it would be intuitively sensible to not act in any sexual way towards someone else unless you’re sure the other person doesn’t mind. Tricky though... how do you know if they like it before you’ve done it? Just do it once then… still not sure? Did they even notice? Maybe it was well-received, and you can ramp it up to the next level? Try again then. And again. So what if you earn a little telling off… or a slap… or loss of friends, family support, your job… so what if you go to prison and the other person goes to therapy… You see how quickly grey areas can escalate. As an additional confounding factor, males are commonly susceptible to the sexual overperception bias1 (SOB) – an exaggerated perception of sexual interest from females.

The evolutionary grounded logic of error management theory2 (EMT) explains such biases in human decision making as adaptive assessments of the costs vs benefits of any given situation. In the case of the SOB, mistakenly not perceiving sexual interest means you’ve just missed a valuable mating opportunity (your genes don’t get passed on!). Mistakenly over perceiving sexual interest may have historically led to a slap, but at least your genetic line remained intact. It’s easy to see how such a bias in male mating strategies developed, but less obvious to determine why this behaviour still occurs. Our goal was to apply this evolutionary logic to the issue of sexual harassment in order to shape future intervention strategies. To do this, we created a survey that measured a variety of individual differences to establish correlations between these and “pro-harassment” scores, or "P-scores" for short.

Here’s what we have so far… 

Firstly, the confusing part, which I shall attempt to explain as I go along: The variance(similar to the range [lowest to highest scores] but based on standard deviations [the average differences between each of the scores) in male P-scores was significantly different to that of females – not only does this suggest that females are more unanimous in their perceptions of harassment behaviours, but also that there’s a much larger portion of males that believe these types of behaviours will effectively help them attain and/or maintain a partner. Ugh, men, right? But hang on… Despite this, in an unexpected turn of events, there was no significant difference overall between how many males and females perceive a higher-than-average-rate of positive outcomes for sexually harassing behaviours. Ugh, women, right? Does that mean we’re not so different after all? Probably not. Because, what's important to consider is the distinct variations in the stories being told by the 2 data sets – yes, everyone equally perceived positive possibilities of sexual harassment, but the individual factors, and possibly the surrounding circumstances that lead to this deviant take are markedly different between the sexes.

The table below displays the noteworthy individual difference that correlate with P-scores for males and females;

 P-score Correlator 
Individual DifferenceMaleFemale
Age Yes
Agreeableness Yes
Hostile SexismYesYes**
Benevolent SexismYesYes**
Sadism Yes
Intrasexual Competitiveness Yes**
Rape Myth AcceptanceYesYes**
Have Been HarassedYes 
Have Harassed Others Yes**

To clarify – each sex has 8 correlational factors (pure coincidence!). Of the male correlational factors, 50% are personality* aspects, innate factors that relate to the self and likely involve little, if any, conscious deliberation to formulate. The majority of female factors, however, are based on their perceptions of others, which would indicate a necessity for socialisation and at least some deliberation to formulate. Females correlations are largely associated with antisocial attitudes** and, in the context of this survey (featuring a specifically male-on-female-harassment vignettes measure), these could be considered anti-women attitudes. This suggests that these females hold negative views of other females and therefore see a positive side to harassment happening to other females (likely at least partly due to sexual competitiveness as research has demonstrated females to engage in passive-aggressive sexual competition strategies3), but not necessarily to themselves, nor carried out by themselves. Thus, whilst females are demonstrably as likely as males to see the pros of sexual harassment, it cannot conclusively be argued that they are as likely as males to engage. When you consider evolutionary theory too, the reasoning behind male unintentional harassment engagement could actually be considered somewhat less antisocial than the reasoning behind female harassment advocating; nonetheless, harassment engagement is ultimately the most unacceptable and more damaging issue of the two.

So, whilst perhaps equally likely to hold pro-harassment attitudes, males and females’ reasons for these, and their subsequent behaviour outcomes are markedly different: Males appear more likely to engage in harassment behaviour and, although arguably females should be supporting each other in such matters, decreasing incidents of harassment from occurring in the first place is the primary matter to be addressed herein.

Examining male pro-harassment factors.

The fact that male P-score correlations were primarily made up of personality factors could fit well with the theory discussed herein. Perhaps narcissism and extraversion = elevated confidence = a logical precursor to elevated sexual overperception. Even factors that within females appear antisocial/anti-women (RMA and sexism) could, from the male perspective, tie-in with an age-old predisposition to believe that females secretly desire to be taken by a man, akin to an excessive overgeneralisation of sexual overperception. But at this point in the discussion, it should be noted that EMT and the SOB are never excuses for harassment behaviour engagement. An evolutionary predisposition is no excuse; our species is capable of using reason to influence the behaviours (yes, even “instinctive” ones4) we choose to engage in. It’s not merely our reproduction-oriented instincts that have evolved, but also more general survival-oriented instincts in the form of pro-social behaviour and advanced learning and adaptive abilities, all of which would draw us away from potentially status-damaging, antisocial behaviours. Our behaviour can be a choice, albeit an often convoluted one.


Choice signpost

How does previous harassment contribute?

Let’s not forget about those previous harassment experience factors. The number of participants who had themselves been harassed previously or who had engaged in harassment previously (and were willing to admit it!) was very low, making it difficult to draw conclusions. However, they are nonetheless interesting themselves. It would seem fairly intuitive that if you’ve previously harassed then you would do it again (as per the female correlation – the male correlation was also high but the low sample-size for this factor may have prevented the results being significant). But one would expect those that have experienced harassment themselves to avoid engaging: Not so for the males. Why? Have they normalised such behaviour, therefore now believing it to be acceptable? Were their harasser’s female and they now engage in harassment as a form of revenge? Did their harassers reap positive outcomes from them, and they want the same results for themselves? Definitely another matter for further investigation; with a larger sample, these may yet prove poignant in future research.

Predicting and preventing sexual harassment.

The final step of this study was to attempt to use the correlational factors to create a predictive model of P-scores. Unsurprisingly (based on previous literature5), rape myth acceptance is predictive of both male and female P-scores. The male model also included benevolent sexism (again supported by previous research6), and extraversion and narcissism (both associated with confidence, thus, as discussed, possible inflation of and/or inclination to act on the SOB). The female models only other addition to rape myth acceptance was agreeableness. Interestingly, extraversion vs. agreeableness could again support the notion of males having a higher likelihood to engage than females: Extraversion vs. agreeableness could indicate aggression vs. passivity, the dominant vs. the submissive. Our next step is to try and encapsulate this very line of thought – do males and females with high P-scores really both have increased likelihoods of engaging in harassment? Or are females with high P-score just more accepting of male-on-female harassment due to their agreeable natures and competitiveness against (or intense dislike of?!) other women? Additionally, there are many other avenues still to investigate beyond biological sex, such as same-sex relationships, culture, harasser and victim relationship etc., any of which could hold divergent primary predictive factors that may differentially shape intervention development. But the evolutionary theory behind the behaviour remains the same and herein arguably remains supported. This theory will ultimately shape the underlying structure of the interventions developed (with the aim to manipulate the perceived costs and benefits of engaging in harassment behaviours), with the primary predictive factors shaping the method in which this achieved (e.g. via the difference in language used to appeal to different personalities). 

Balancing decisions

To realise this goal, our current study seeks to refine the vignettes to generate more concrete answers to our first questions; does a pro-harassment attitude indicate an increased likelihood to engage in harassment behaviour (based on ratings of personal beneficial/detrimental effects of potential positive/negative outcomes of harassment behaviour engagement), and does this additional information effect and/or clarify the sex differences observed? Then, with the use of the most predictive factors from within the models created we shall begin to frame educational intervention(s) not just for those deemed potential perpetrators (this isn’t Minority Report after all!), but for everyone, with the eventual aim to introduce a programme within secondary school education. The predictive model will allow selection of a “most-likely to engage” population to trial such an intervention on; if pro-harassment attitudes can be reduced in these individuals, or even in a previous offender sample, then it can be taken to the general public with confidence. By analysing the potential for perpetration and thusly educating those that do not want/mean to become offenders, we can create a proactive approach rather than another attempt at a “treatment” pathway for victims. Reduce the number of offenders and you reduce the number of victims. I’m sure any victim would value never have become a victim in the first place over having an effective therapy.

The journey continues. Please take part below. Your participation will be greatly appreciated.

Survey Link


QR Code survey link



Add a comment