1

In their newest video the "FCKH8" company claim that:

1 out of 4 women will be the victim of domestic violence

Is this indeed the case for women in the US?

George Chalhoub
  • 30,246
  • 14
  • 129
  • 136
SIMEL
  • 29,037
  • 14
  • 123
  • 139
  • 1
    "Skeptics Stack Exchange is for challenging unreferenced notable claims, pseudoscience and biased results." Providing an answer with a question strikes me as *proselytizing*, not challenging. – gnometorule Dec 17 '14 at 00:41
  • 2
    @gnometorule, answering your own question is acceptable and encouraged, see [our meta](http://meta.skeptics.stackexchange.com/questions/1983/should-i-ask-a-question-about-a-notable-myth-that-i-already-know-the-answer-to) and [SE meta](http://meta.stackexchange.com/questions/17845/etiquette-for-answering-your-own-question/17847#17847). There is no reason to let others do research that you have already done. Specifically in this case, I look forward to read other sources and answers, if you have other credible sources that point to a different conclusion, please post them as an answer. – SIMEL Dec 17 '14 at 07:13
  • @IlyaMelamed - given that pretty much most recent research (even given that much of it is suppressed for political reason) points that domestic violence is roughly on par between genders as victims ([example](http://www.saveservices.org/2012/02/cdc-study-more-men-than-women-victims-of-partner-abuse/)), this does seem like proselytizing, since you singled out women as if they are the only victims, blindly following FCKH8's propaganda. – user5341 Dec 17 '14 at 18:19
  • @DVK, I didn't chose to focus on the statistics for Women, FCKH8 did, I just wanted to check their claim. The example you give is citing the 2010 CDC survey that appears in my answer. The sources I provided in my answer address the statistics for men as well as women, in the quotes I took from both pf them I cited the statistics for men as well. If you feel that more attention should be given to the figures for men in the answer, you can edit it. If you know of other sources that claim differently, please add them in a new answer. – SIMEL Dec 18 '14 at 08:18
  • @DVK, Also please note that the example that you give addresses a different statistic than the one in the video. The statistic you are referring to is people who suffered domestic abuse in the last (2010) year, the statistic that the video addresses in suffering domestic abuse during a lifetime, as Thales pointed out in a comment on the answer this figure is affected by the norms of past years and decades. Perhaps this is an issue that should be addressed an answer as well. – SIMEL Dec 18 '14 at 08:22

1 Answers1

4

I've reached out to FCKH8 asking for their sources for the figure, sadly, they haven't answered me yet (~ 2 weeks at the time of writing).

The figures in different publications are around the 25% figure for women, and this figure is confirmed by studies.

There are many sources citing the 1 in 4 statistic. It seems that the source for many of those claims is the survey Extent, Nature, and Consequences of Intimate Partner Violence: Findings From the National Violence Against Women Survey, published by the Department of Justice, which states in its abstract:

Among the survey findings are that intimate partner violence is pervasive in U.S. society, with nearly 25 percent of surveyed women and 7.6 percent of surveyed men reporting that they were raped and/or physically assaulted by a current or former spouse, cohabiting partner, or date at some time in their lifetime; 1.5 percent of surveyed women and 0.9 percent of surveyed men said the were raped and/or physically assaulted by a partner in the previous 12 months. Based on these estimates, this report indicates that approximately 1.5 million women and 834,732 men are raped and/or physically assaulted by an intimate partner annually in the United States.

The survey was published in the year 2000 and it analyzes data from November 1995 to May 1996, so its conclusions are about data which is almost 20 years old. Nonetheless it's cited in many credible places: The American Bar Association, The CDC, The National Coalition against Domestic Violence and others. The CDC used this survey as a source on their 2003 study Costs of Intimate Partner Violence Against Women in the United Stated.

However, there are some sources which claim a different figure, one example is the CDC itself, who, in their 2010 The National Intimate Partner and Sexual Violence Survey summary report put a higher figure:

More than 1 in 3 women (35.6%) and more than 1 in 4 men (28.5%) in the United States have experienced rape, physical violence, and/or stalking by an intimate partner in their lifetime.

[...]

About 1 in 4 women (24.3%) and 1 in 7 men (13.8%) have experienced severe physical violence by an intimate partner (e.g., hit with a fist or something hard, beaten, slammed against something) at some point in their lifetime.

The second figure is closer to the original 1 in 4 figure.The two sources do have different definitions of domestic abuse; the 2000 survey talks about rape and/or physical assault while the 2010 source also adds stalking. Even Though some of the sources that cite the 2000 study add stalking to their definition of domestic abuse. The 2000 survey does touch and expand on stalking in its body.

Community
  • 1
SIMEL
  • 29,037
  • 14
  • 123
  • 139
  • _In their lifetime_ is a really dangerous criteria. If someone suffered this kind of violence 40 years ago, that number would still show today. Violence against partners change during time periods, and research like those don't seen to... adjust to this peculiarity. – T. Sar Dec 17 '14 at 10:19
  • @Thales, I didn't choose which statistic to use. If you go over the sources you can also read statistics of how many people suffer each year. – SIMEL Dec 17 '14 at 10:37
  • 1
    Oh, I'm not saying that this is your fault. What I'm saying is that this kind of study can be dangerous if it is not taken with a really criptic eye. The U.S. Is a _huge_ country. The differences between different zones can be really significative. Without narrowing criteria, simple surveys like those have almost nill value. – T. Sar Dec 17 '14 at 10:41
  • That last study, for example. 16k people can give you some really biased results depending on where the subject lives or the average family income. You can't really consider this a _random sampling_ for a country with more than 300 _million_ people. This is a sample so small that don't really make any sense on the bigger picture. – T. Sar Dec 17 '14 at 10:47
  • @ThalesPereira I'm not a statistician. Doesn't http://en.wikipedia.org/wiki/Sample_size_determination#Estimating_proportions_and_means imply that the sample size (e.g. 16,000) required for a given accuracy is independent of the population size (e.g. 300,000,000)? It seems to me that 16,000 is a **HUGE** sample size. The study may have problems (e.g. it's only for people who answer the phone, only for people who are willing to discuss it over the phone, no trained counsellor in-person to help ensure that people use the standard definition of "assault" etc. when they report their experience) ... – ChrisW Dec 17 '14 at 11:11
  • ... but the size of the sample isn't one of these valid problem. – ChrisW Dec 17 '14 at 11:11
  • 1
    @ChrisW That estimated sample size is really basic and is valid for the same _population_ (i.e.: more or less homogeneous caracteristics, like a small town) but not for something that have wildly differences. On something with around 300 million people, you don't have _a_ population, you have several different populations with unique traits that must be respected. This kind of strategy for a study is why most Brazillian statistics fail miserably. – T. Sar Dec 17 '14 at 11:18
  • @ChrisW For a example, last election for Governor on the state which I live sampled around 10k people, giving the victory for the candidate that lost. The study showed something like 51%/49%. Not only the study winner did _lost_, but by a extremmely _huge_ margin, 61% against 39%. – T. Sar Dec 17 '14 at 11:24
  • @ThalesPereira My guess is that a 16K sample that is all taken from the same small town would be a very bad predictor of the national population, but that a 16K sample taken by truly-random selection from accross the nation would be a good indicator. So what matters is whether they were selected randomly (therefore well-distributed across the population). Imagine there are isolated pockets of rich people and of poor people: either 16K random samples will find (sample from) those pockets, or the pockets are so small (e.g. less than one in a thousand) as to be insignificant. – ChrisW Dec 17 '14 at 11:25
  • @ThalesPereira There are lots of reasons why polls might be inaccurate. It's theoretically possible to poll everyone in (100% of) the population and still get a different actual result on election day. – ChrisW Dec 17 '14 at 11:27
  • @ChrisW If you want to do a sampling to gauge national matters, you should also provide several other criteria so the statistic can be valid for that population. Think about income - if you sampled 16k random people, there are huge chances that every person on that sample is poor just because there are way more poor people than people on other income strips. Statistics are _really_ tricky. – T. Sar Dec 17 '14 at 11:29
  • Let us [continue this discussion in chat](http://chat.stackexchange.com/rooms/19546/discussion-between-chrisw-and-thales-pereira). – ChrisW Dec 17 '14 at 11:36