In part 1 & part 2 it was explained that In September of 2014 in the Journal of the American Board of Family Medicine an article was published titled “Characteristics of Men Who Perpetuate Intimate Partner Violence.” The article, as so many others before it, focused solely on men as perpetrators and women as victims. It estimated that 1 in 5 men admitted to being violent toward their spouse. The media caught wind of this and a flood of articles were published with the headline “1 in 5 men admit to violence toward spouse.”
A public database was used for this research. I asked the researcher for the raw numbers for females admitting violence and he refused saying I needed to find a statistician to help me obtain that data. That smelled a little stinky to me and it made me wonder if he had something to hide. I went about figuring a way to get the data myself and was pleasantly surprised to find that it was open for anyone to see and was on-line to boot!
I taught myself the basics to get to this raw data and first looked into the ways the data was collected. They used two questions which were drawn from a nationally representative database. (NCS-R) One of the two questions asked about the respondent’s usage of minor violence towards their spouse and the other asked about severe violence. The responses were broken down into four possibilities which detailed how often the behaviors occurred:
These categories gave one a sense of the frequency of the behaviors being studied. But here is the kicker. The researchers didn’t use these four responses even though they were available on the database. Here’s a quote from the research paper that describes what they didi:
In brief, the dependent variable IPV perpetration was assessed by asking: “Over the course of your relationship, how often have you ever done any of these things (pushed, grabbed, or shoved; threw something; slapped or hit; kicked, bit, or hit with a fist; beat up; choked; burned or scalded; threatened with a knife or gun) to your current spouse/partner?” Responses included often, sometimes, rarely, or never. We dichotomized responses into any/none.
In other words, by “dichotomized” they mean they turned all the different four responses into either “yes” I committed violence or “no” I did not commit violence. They took any answer that was not “never ” as constituting an incident of domestic violence. With no way of interpreting the frequency of these behaviors we are left just guessing unnecessarily. This limits the usefulness of the data. The chart below gives you a quick look at all the answers that were not “Never.” It shows all of the positive responses (the admissions of violence) to the two questions from the database where respondents answered “Rarely”, “Sometimes” and “Often.” Notice that 87% of these responses were “Rarely.” Knowing that the vast majority have answered rarely puts a very different spin on the data. But since the study has removed this information it leaves the reader unaware of any frequency information and it is anyone’s guess what people will assume. If you only read their study with their dichotomized data and don’t know about this detail of the data you might assume that all of those responses were incidents of serious violence. Have a look at this chart and see how the vast majority of answers were “Rarely” (457) and there were very few “Sometimes” (63) and fewer still “Often.” (7)
Why would researchers do this sort of thing? I am not sure what their reasons were but it is clear that by counting the incidences as they did it will tend to inflate the appearance of domestic violence. This gave them the ability to make the claim that 1 in five men “admitted” to being violent towards their spouse. Just imagine if they had not dumbed down the data. They would have had to say “One in five men admitted to rarely being violent towards his spouse and one in 1000 answered “Often.’ It just loses its sexiness doesn’t it?
Then the question arises why would any researcher want to diminish the information in his data? It might have been very instructive if they could differentiate the different levels of frequency of violence. They could then say things like “Those men who claimed to “often” use violence towards their spouse were more likely to x than the men who said “rarely.” This could be very helpful information to clinicians, law enforcement, and many others but we simply don’t see that level of detail since the data has been “dichotomized.” My guess is that the motive here is to inflate the appearance of domestic violence and by doing this they get more likelihood of funding for their next study. But this is just my guess.
When people think about domestic violence they are often thinking of someone being severely beaten. They are not thinking of someone who gave a gentle push or grabbed an arm in a moment of irritation and both parties then calming down shortly thereafter. But the way these questions were asked all of the “rarely” responses could be just that: a momentary irritation. One of the questions asked a list of behaviors including if you had ever pushed or grabbed your spouse. If you pushed your spouse 20 years ago and never pushed her again you would answer yes to this question and would be counted as someone who admitted to violence in relationship. The way the questions were worded leaves us wondering about the severity of violence associated with the “Rarely” responses. It is possible that with the wording of the questions that the “Rarely” category might be a slight push every twenty years. So just to experiment, let’s exclude these “rarely ” responses and only count the “sometimes” and “often” responses as being evidence of more serious domestic violence the situation changes dramatically. Now instead of being 1 in 5 it is more like 1 in 50. Even that I think is not accurate. If you exclude the sometimes responses and only count for the question about severe violence the figure drops to 1 fifth of one percent .17% (about 1 in 500) Very very low but these researchers tried to paint a picture using all of the positive responses as being a “yes” thus creating the appearance of a more widespread problem.
But with these caveats let’s accept this as it is and move on.
We have seen how this researcher harbors ideas that are likely to diminish the chances of male victims being highlighted. We have seen how the data was “dichotomized” and how this may have altered the meaning of the numbers to the general public. Now let’s turn to the stunning fact that the database he used for this study to show how 1 in 5 men admitted to being violent with their spouse actually showed that women admitted to more violence in relationship than did the men, sometimes by as much as double. Let’s look at each of the two questions.
Here’s the first:
MR42. F (RB, PG 56)
People handle disagreements in many different ways. Over the course of your relationship, how often have you ever done any of these things on List A to your [(current)] [(spouse/partner)] – often, sometimes, rarely or never?
· Pushed, grabbed or shoved
· Threw something
· Slapped, hit, or spanked
Let’s have a look at a chart that shows both men’s and women’s response to that question.
Note that the majority of responses were “Never” with “Rarely” coming in a distant second. Then note that the “Sometimes” and “Often” responses are a very small number in comparison. You will see that of the responses that admitted to any violence (rarely, sometimes, and often) the female totals were always higher than the males. In the sometimes and often responses they were almost double. This is remarkable but it got buried by the researchers only focusing on male violence. Also note that the males admitting to minor violence are about 15.5% of the total while the females admitting minor violence are about 21%. That is quite a gap.
So we can easily see that the researcher simply ignored the female data. It was there but he chose to turn his head.
Next up is the question about severe violence. Here is the question as it was asked:
MR44. F (RB, PG 56)
Now looking at List B, over the course of your relationship, how often have you ever done any of the things on List B to your [(spouse/partner)] – often, sometimes, rarely, or never?
Kicked, bit or hit with a fist
Burned or scalded
Threatened with a knife or gun
See the chart below and notice that the same patterns play out in this chart with the major difference being that the numbers are sharply diminished. Again notice that the female numbers are always higher than the males and in the “sometimes” and “often” responses are double or more.
This seems like a very important difference that is contrary to the stereotype that has become the norm. The least that needs to be done is for the researchers to attempt to explain this difference. I am willing to bet that their explanation would focus on the man’s unwillingness to tell the truth. This explanation might have some credibility since men are far more likely to face harsh judgement and shaming for admitting hitting a woman while women do not face nearly the same sorts of judgements for hitting men. But the data does not support this idea. There were other questions on this same database about domestic violence and one of those asked the respondent for the frequency of how often the spouse hit them. If we assume that men were lying about their violence we would expect that the women’s responses to how often their spouse was violent towards them would show that their masculine partners were more violent and the women’s numbers about the men being violent would be greater than the men’s numbers. But that is not what the responses show. The responses show that women reported that men hit them less than the men report the women hitting them. This seems to support the idea that women are more violent in relationships (at least in this sample) just as the raw data from these questions suggests.
It is also worth noting that just as the researchers “dichotomized” the Rarely, Sometimes and Often responses into yes or no, they have also combined the question about severe violence and minor violence into one unit that is expressed as a yes or no. If someone answered affirmatively to either of these questions it was counted as an incident of violence. But keep in mind that there were nearly seven times as many affirmative responses to the question about minor violence when compared to the severe violence. These important differences disappear when the data is simply totaled and you ignore both the frequency and the severity. Again, the same theme plays out that “dichotomizing” the data and now the questions puts strong and unnecessary limits on its usefulness. The only reasons I can imagine they would want to do this would be to inflate the appearance of domestic violence. Just as the activists, media and so many others try to paint an exaggerated picture we now see the researchers apparently taking a similar path.
It seems to me that List B is more representative of what most of us consider domestic violence. Kicking, beating up, choking, threatening with knife or gun etc. These are indicators of serious violence. If we only look at the percentages of this question we see that the number of females admitting severe violence totaled 3.1% (approx. 1 in 32) while the males admitting severe violence totaled 2.2% (approx. 1 in 45). That says that nearly 60% of those admitting to severe violence are women. What? Has anyone heard any research that points to those numbers? No. And that is the point of this article. We have heard only half the story and as evidenced by this research the numbers were there, the researchers simply opted to ignore them thus leaving most of us in the dark about the realities of domestic violence.
We have seen how the ideas and attitudes of the researcher played out in only reporting one side of this story. We have seen how the “dichotomizing” of the data and the questions basically dumbed down the data and made if less useful by making it a simple yes or no. We have seen how very shocking and informative data that conclusively shows that women admitted to being more violent in relationship was ignored and unreported. This all facilitates the promotion of the default narrative of women as victims and men as perpetrators by only telling the story about male perpetrators and female victims. We have seen how this works and the powerful national media’s willingness to promote this half story on a national level.
Look at the headline below. Now you know this headline should actually read “1 in 4 American Women admit to domestic violence.”
Can you imagine seeing an article like the one pictured below in a mainstream media publication? I would bet not. But like it or not, that is actually the truth.
It’s time we started holding researchers, the media and all of those connected to domestic violence accountable. This charade has gone on far too long.
Tom Golden, LCSW is an author and psychotherapist who has been writing on issues concerning men, boys, and gynocentrism for many years. Tom has written three books, Swallowed by a Snake: The Gift of the Masculine Side of Healing, The Way Men Heal, and Understanding the Unique World of Boys. He has a blog at Menaregood.com and also has a youtube channel menaregood.