Mumsnet rape and sexual assault survey results, and a word on statistics

// 12 March 2012

Tags: , ,

Mumsnet we believe you campaign logo.gifOn Friday, Mumsnet launched its “We Believe You” campaign, aimed at highlighting the prevalence of rape and sexual assault in women’s lives and busting the myths surrounding rape. The site has also released the results of its survey, in which over 1600 site members participated. The results won’t come as a surprise to anyone here:

The survey shows that, of respondents:

One in 10 (10%) had been raped

Over one-third (35%) had been sexually assaulted

Almost one-quarter (23%) reported being raped or sexually assaulted four or more times

In two-thirds (66%) of cases the women knew the person responsible

Many women felt unable to report rape or sexual assault:

Over four-fifths (83%) of respondents who had been raped or sexually assaulted did not make a report to the police

Over one-quarter (29%) didn’t tell anyone at all, including friends or family, about the assault/rape

Over two-thirds (68%) said they would hesitate reporting to the police due to low conviction rates

And over half (53%) would not report due to embarrassment or shame

The results also reveal that most women feel that rape victims are treated poorly:

Nearly three-quarters (70%) of respondents feel the media is unsympathetic to women who report rape

Over half (53%) feel the legal system is unsympathetic

And over half (55%) feel society at large is unsympathetic

I think it’s fantastic Mumsnet are engaging with this issue, and I love the name of the campaign, but I do think the results relating to the number of women who have been raped or sexually assaulted need to be used carefully.

In the Guardian, Mumsnet co-founder Justine Roberts says: “We simply shouldn’t accept that we live in a country where one in 10 women are raped and over one third sexually assaulted”. I agree with the sentiment, but the Mumsnet survey was self-selecting, meaning the results cannot simply be extrapolated to the wider population. For example, it may be that women who have experienced rape or sexual assault would have been more likely to respond to the survey than those who haven’t. So all the survey can actually tell us is that one in ten of the women who completed the survey have been raped, not one in ten of all British women.

While the above in no way undermines the significance of the data relating to the reporting of these crimes, and while the data does contribute to the overall picture of women’s experiences, rape apologists and anti-women commentators love to question the validity of statistics in order to try and undermine work to tackle the issues surrounding sexual violence, so it’s important that we use them correctly.

It’s difficult to get an accurate picture of the exact rates of sexual violence against women. Truth About Rape has more information here, including British Crime Survey data showing that one in ten women have experienced some form of sexual victimisation, and one in twenty have experienced rape. The BCS is generally thought to paint a more accurate picture of crime rates than the recorded crime figures, as it asks people whether they have been a victim of crime, regardless of whether they reported it to the police. (Of course, women would not necessarily feel comfortable disclosing this information to a researcher, so these figures can’t be fully relied upon either.) Other studies have found that one in four women experience rape or attempted rape.

What’s important here is the fact that rape and sexual assault are experienced by a huge number of women, be it one in twenty or one in four, and the need to tackle it is no less urgent, whichever statistic we choose to use. Let’s just cover our backs and make sure we use them correctly.

Comments From You

Helen G // Posted 12 March 2012 at 2:14 pm

Great post on an important subject. This survey is a start, a solid start, and it would be good to see other groups and organisations take it and expand on it further so we can get an even more accurate picture to help us all work out meaningful, integrated and compassionate ways to reduce the incidences of these horrific crimes.

Feminist Avatar // Posted 12 March 2012 at 2:17 pm

On the issue of statistical validity, the survey participants may be self selecting but that doesn’t mean we can’t extrapolate from it- we just need to figure out what group in represents (surveys that choose their populations do so to ensure broad representation). 1600 is a massive survey, and anything above 30 has statistical validity. What we need to know is who these women are- are they all white or aged between 18-24 for example, or are they reflective of the broader population? I guess we can say it is likely they are all mothers and most are likely to be childbearing age. It’s a shame that this info isn’t part of the analysis (if it was collected), because it would help give force to the numbers.

Laura // Posted 12 March 2012 at 2:42 pm

Thanks for clarifying that, Feminist Avatar. I filled out the survey and don’t recall being asked for any additional personal information, other than perhaps age.

Philippa Willitts // Posted 12 March 2012 at 2:56 pm

Thanks for this, Laura. As far as I’m concerned then if it’s happening to one woman then it’s a crisis, and it’s urgent, and we need to act. The figures are important, especially for things like fundraising for Rape Crisis centres, and being taken seriously, but the human cost is vast, whichever of the stats are correct.

Rose // Posted 12 March 2012 at 4:09 pm

… Sure, in a way women with bad expirences were more likely to fill in the form… but women with PTSD with a related cause may have avoided the survey as not to bring up those memories. Theres also the issue of only women who have come to terms with the fact that ‘that’ happened will report having been raped. (Look at the work of Diane EH Russell on rape in marriage).

There will always be problems with surveys, but from what I’ve seen and heard of the world these are the most convincing stats yet.

Good post.

Laurel // Posted 12 March 2012 at 4:37 pm

i think one reason for variance might include peoples definitions of rape, both legal and personal. what feels violating or what feels like rape can vary. most people in feminist circles wont limit rape to just PIV. It could be applied to any sustained sexual contact where one was violated against their will plus any entering of body parts seen as sexual, or some people will say anything that the survivor wishes to define as rape, which would actually phase out a lot of things that we may consider rape, such as pressure or bribery etc.

Shadow // Posted 12 March 2012 at 7:59 pm

Unfortunately survey did not ask respondent to name sex of perpetrator(s) but despite this we all know 99.9% of those who commit sexual violence against women and girls are male. So this means it is males who are the ones committing sexual violence against women not those non-defined person(s).

What is important is that many many males of all ages are continuing to commit sexual violence against women and girls and their likelihood of being prosecuted let alone convicted for the crime of sexual violence against a female/females is almost zero. That is the reality of living in a male supremacist system.

Sue Marsh // Posted 13 March 2012 at 9:42 am

Psephologist would call this a “voodoo poll” but considering the subject matter I’m not so sure. Mumsnet has a vast readership. 1,600 responses is a very significant sample – on a par with national opinion polls

People who use Mumsnet feel safe – probably safer to respond to this survey than others. The stats it threw up are not so wildly out of kilter with other figures on rape/abuse.

To stop this from being judged, Mumsnet, might want to select a further sample at random, hopefully taking people from all ethnic groups, age groups, income brackets etc and ask the questions again. It would be incredibly interesting if the results were broadly the same.

Unity // Posted 14 March 2012 at 1:58 am

Regarding the statistical validity of the poll there are some issues, particularly around demographics and selection effects.

Age is a particularly important factor here, in part because different age groups are subject to different levels of risk but also because trends change over time – the annual prevalence of rape has declined to a degree over the last 20-30 years and this feed through into estimates of lifetime risk in terms of cohort effects, i.e, the annual risk for a 25 yr old in 2012 is a bit lower than it was for a 25 yr old in 1972 so when you ask about lifetime experiences, responses from older women will skew the figures upwards and provide a somewhat false picture of the current situation.

That doesn’t that that lifetime risk figures aren’t important or useful for other purposes but they do need to handled with a degree of care.

Also, to answer Laurel’s point, evidence from the most recent BCS survey, in which they’ve trialed an alternate set of questions in the intimate violence module, appear to indicate that the prevalence figures go up, not down, if the question allows women do define rape in their own terms rather than asking them whether their experiences fit a specific legal definition of rape, so its unlikely that that’s a factor in this case.

As things stand, extrapolating from this survey is rather risky without access to key demographics – it works as an attention grabber but trying to make anything more of it than that will run into the usual problems, i.e. wrangling over statistic validity, etc, which deflect attention away from the key message, which is simply that its a hell of big problem no matter what kind of numbers these surveys generate.

However, for the record, the 1 in 4 statistic for rape doesn’t stack up – the source data isn’t comprehensive enough to generalise onto the wider female population and the figures one which this is based are now 20-25 years old and rather out of date.

1 in 4 is, however, pretty much on the money for domestic violence.

The BCS estimate of 4-5% is age dependant – the question they use asks whether respondents have been raped since the age of 16 and doesn’t account for future risk, so it shouldn’t be used as an estimate of lifetime risk.

However, if you model things properly from current annual prevalence data then current individual lifetime risk around 9% for women over 16 and 10.5% (including under 16s) – this is after you adjust for repeat victimisation, and if you weight that figure for age and trends in prevalence rates you’re looking at around 14-14.5% for current lifetime prevalence, so anything in the 1 in 7 to 1 in 9 ballpark is a pretty good shout.

What really doesn’t get enough attention in all this, because its quite difficult to model from existing statistics, is the extent of repeat victimisation. For women who have been raped, the risk of being raped more than once is somewhere between 1 in 4 and 1 in 3 and, as you might well expect, the main reason that figure is so high is because rape is a pretty significant part of the overall landscape of domestic violence.

If you want a figure that busts through some of the myths about stranger rape then the repeat victimisation estimates are the one’s to push hard, as these are rarely, if ever, reported in the media.

Have Your say

To comment, you must be registered with The F-Word. Not a member? Register. Already a member? Use the sign in button below

Sign in to the F-Word

Further Reading

Has The F-Word whet your appetite? Check out our Resources section, for listings of feminist blogs, campaigns, feminist networks in the UK, mailing lists, international and national websites and charities of interest.

Write for us!

Got something to say? Something to review? News to discuss? Well we want to hear from you! Click here for more info

  • The F-Word on Twitter
  • The F-Word on Facebook
  • Our XML Feeds