Rape statistics.

Here at my college, I often see signs on the wall, courtesy of the Center for Womyn’s Concerns, the feminist group on campus, with the statistic: “45% of all women who are raped tell NO ONE.” The figure may not be the right one, but what I’m wondering is this: How is such a statistic gathered? If they told NO ONE, (The sign emphasizes “NO ONE.”) then how can we presume to know the percentage?

Probably goes like this.

Guy says ‘I had sex with Mary’

Someone asks Mary ‘Did you have sex with that guy?’

She says ‘No, of course not!’

They figure if he says he had sex with her and she says he didn’t, it must have been while she was unconscious! She was raped!!!

You could also make a pretty good case that many of that 45% are murdered, and can’t tell anyone.

I can answer this. Confidenti surveys of abuse victims often specifically include questions about who, if anyone, they told about the experience. So what you saw is an estimate based on psychological or sociological research.

Iamthewalrus, I don’t think so.

Most rapes are what are what are now known as “date rapes.” So women don’t tell because they:

  1. Don’t think anyone will believe them.
  2. Think they will be told is was their fault.
  3. Were raped by a family member or friend of the family and fear what the effect will be if they tell.

Among many other reasons.

I knew that, I was just being silly.

A lot of that has to do with new definitions of rape. Some of these surveys consider it rape if the woman has no memory of the actual sexual encounter, or even if she was just drunk at the time. Some consider it rape if the woman was coerced into it - not threatened with physical harm (which I would consider rape myself) but told they would be dumped, or pressured to as part of some kind of initiation. I think these new definitions are a bad thing, as they reduce the severity of REAL rape (forced sex or sex with an unconscious person).

Keep in mind, though, that a lot of women will not reveal that they’ve been raped, regardless of the confidentiality involved. This is most often true of the most common kinds of rape- those involving aquaintences and dates as well as family members.

It makes sense that a lot of rape victims wouldn’t tell anyone out of a sense of shame or denial. At the same time, I agree that it would be pretty difficult to measure the percentage. Presumably, a lot of rape victims wouldn’t even admit it on a confidential survey.

None of these responses seems to address the question, though. The statistic is garnered by a survey of abuse victims, that would mean the victim sought out professional help, which most certainly does count as “telling someone.” Perhaps they meant “told someone besides a person in a professional context,” but that just doesn’t make any sense to me. It seems like they are pointing out the frequency of unreported rapes, which would be difficult to number.

Ditto Tzel. This thread is not really about rape, it’s about statistics. Or, more specifically, this thread is about unreported statistics.

For example, check out this quote from the U.S. Census Bureau, at http://rates.census.gov/ (when you get there, click on “Census 2000 Final Response Rates”, at the top):

I find the above to be very difficult to understand. It sounds like 67% of all Americans responded to the Census questionnaires. To come up with that percentage, they’d need to know how many people responded to the Census (that’s easy - just count them up) and also how many people did not respond. But how the heck can they know how many people did NOT respond???

Obviously, professional statisticians have a way to estimate how many Americans did not report to the Census Bureau and how many victims did not report being raped.

But to Tzel and I, these are not estimates, but guesses. We’d like to know how these figures are arrived at.

It is particularly annoying to see the Census Bureau (see quote above) proudly announce that the 2000 Census reached 2% more Americans than the 1990 Census did. But if both figures are just guesses, then it’s all worthless. For all we know, the 2000 Census actually reached fewer people thant the 1990.

Getting back to rape, the same complaint applies. Suppose next year more people report being raped, but the statistians may decide that actally fewer rapes occurred, but more of them got reported. Come on, where do they get these numbers from?


“There are three kinds of lies: lies, damned lies, and statistics.” - Benjamin Disraeli

Sounds simple to me. They know how many forms they sent out. The percentages given are the percent that were sent back.

Regarding the OP, I agree that the statistic is probably bogus. But the methodology is based on sampling. If you ask a woman if she ever told someone about being raped and she says no, she has obviously told you about it. But the assumption is that for the vast majority of people who have not been approached by survey takers, it will continue to be true that 45% have told no one.

Um, Izzy. The census bureau DOES know how many FORMS they sent out. They DON’T know
A. how many forms went to homes where no one lived vs. no one answered.
B. How many forms went to homes that were possibly second homes etc.
C. etc. etc. etc.

wring,

The response rate just means that xx% responded. Sure some didn’t respond because of the factors that you mentioned. I didn’t see anyone claim otherwise.

The census figure is relatively easy to estimate. We have good sources for the total population, including hospitals, employment, school, voter registration, etc.

Rape statistics are more difficult. There is no rational way to estimate or even guess the number of rape victims who literally tell no one. We can certainly estimate the number who do not tell the police (IIRC it’s a fairly high number).

Of course, you forget that any time a woman has sex with a man it’s rape. :slight_smile:

I don’t doubt that there is a statistical method to determine unreported rapes, but I do doubt the source in this case.

The definition of rape is highly variable, and a statistic without a definition is useless.

A national census is not a very reliable or efficient way of compiling population statistics (although it does have its uses in the compilation of some other statistics). Reliable figures can only be achieved by correcting the raw data on the basis of targeted sampling. They know how many houses are probably empty because they send people round to some of them to check. The decision by the U.S. Census Bureau to introduce those techniques caused some controversy as the results had implications for the allocation of federal budgets.

Regarding the OP, cher3 seemed to have the correct answer. Ticking a box on an anonymous survey is not exactly telling someone, and those findings can then be generalised. Whether the resulting statistic is likely to be accurate is anyone’s guess. However, the suggestion that almost half of all women who are raped do not discuss their experiences with anyone is not implausible. The shock value of the slogan, for all its spurious precision, therefore makes a legitimate point.

They get rape statistics from anonymous surveys. They ask the question “Have you been raped?” and then ask "Did you report it to the police?, Did you tell you family?, etc. then they match up the results.
However, the statistic you mentioned is probably bogus. What advocacy groups do is to mess with the definition of rape so that it includes having sex while inebriated or having sex with someone because you were tired of them asking. Thus many of those who are classified as having been raped never though of it as rape.
It is true that rape is the most underreported crime but 45% is way too high.

What a fascinating way to make a point: levy a statistic that is designed to frighten and alarm, and deem any attempt to verify it as impossible.

While I am certainly not, and in no way, suggesting that rape is not a serious issue that merits serious consideration, it hurts the cause to flaunt such obviously contrived information. It makes the heralder look like a fool.

One of the striking ways this is illustrated: if there are 10 cats and there are 5 reported meows, can we assume that 50% of the cats meowed?

It’s not true that most research surveys “mess with the definition of rape.” There is a standard measure used–the Sexual Experiences Survey–which requires that the person have been forced or threatened or been given drugs or alcohol. All the questions clearly ask about UNWANTED sex.

Now, I suppose the respondents could interpret this broadly if they wanted to, but it’s not the researchers who are doing it.