Research and devilment | HBR Talk 139


Last week on HBR Talk we discussed some aspects of media manipulation of political discourse among the public. This week we’re going to revisit similar deception engaged in by academic activists who use manipulation of data to promote narratives that it doesn’t actually support. While legitimate research seeks to find solid evidence on which to draw rational, well-supported conclusions, ideologues view research as a tool for confirming their preconceived notions. The same methods and processes legitimate researchers use to weed out unfounded beliefs or unsupported suppositions, ideologues twist to create false support for beliefs they want to maintain, or suppositions they want to believe.

It’s not just researchers. Interdisciplinary studies students do this as well. In our ad for the stream this week I gave the example of reddit feminists I encountered in 2013, using labor bureau data on workplace deaths to promote a female victim narrative. They claimed that in 2012, homicide was the leading cause of workplace death for women and most were killed by their intimate partners. The claim was sometimes stated in response to discussion of men’s workplace deaths, and sometimes in response to discussion of gender symmetry in partner violence. It was used in both instances to imply that women are at higher risk than men, and men are at fault for this. Evaluation of the data they referenced didn’t support their claims or the narrative they were using them to promote. 

They had ignored the vast, overwhelming majority of workplace deaths, which happen to men. By comparing the overall category of homicide to the subcategories of various types of accidents, they falsely presented homicide as a leading cause of women’s workplace deaths, when in fact if you compare category to category and subcategory to subcategory, accidents were far more prevalent. They also ignored the fact that 79% of workplace homicide victims for that year were men. They even ignored the majority of women’s deaths by homicides in the workplace, in which “other assailant,” is the largest perpetrator subcategory and “robber” is the second (both categories in which male deaths also outnumber female deaths,) with the category that includes domestic partner homicides coming in third out of six perpetrator categories, and that didn’t attribute all of the deaths to domestic partners, but to “relative or domestic partner.” More female workers died that year from “roadway incidents” and from “slips, trips, and falls” than homicide committed by “family or intimate partner,” and approximately as many female workers died as a result of contact with objects or equipment as from family-member or parter-perpetrated homicide. In that specific subcategory of workplace homicide, there was a grand total of ten more female victims than there were male victims. 

So… the leading causes of workplace deaths that year killed far more men than women, but if you ignore that, the leading causes of women’s workplace deaths in 2012 were accidents, but if you ignore that, it’s homicides, mostly not caused by family members, but if you ignore unknown killers and robbers, then it’s family members… and if you ignore family members who are not domestic partners, then it’s domestic partners. It’s an epidemic, I tell you! Misogyny!

Why do I suddenly feel like I’ve slipped into the dialogue in a Marx Brothers movie?

Another example I experienced more recently resulted from an argument with the illustrious Professional academic, Michael Flood. I had said, “The biggest problem with feminist research on abuse is double standards. I never trust just the abstract or conclusions from a feminist-led study, because it doesn't show whether the same behaviors were treated differently depending on which sex engaged in or experienced them.” In defending his colleagues against that allegation, Flood posted a link to a study on male victims of coercive control. 

Is It Coercive Controlling Violence? A Cross-Sectional Domestic Violence and Abuse Survey of Men Attending General Practice in England, by Hester, Jones, Fahmy and Feder, had a sample size and strategy that Flood held up as his preferred methods, guaranteeing more accurate and widely applicable results.

Did the work live up to his marketing attempt?

Well… the devil is in the details, isn’t it?

As the title indicates, the method used was a survey given to men who visited doctor’s offices in the UK. The researchers’ definition of coercive control included the experience of a wide list of behaviors, ranging from physical and sexual abuse, to threats, financial limitations, haranguing, and examples of isolation, to speeding with the partner in the car, allegations of cheating, and blaming one’s drinking, drug use, or suicide attempt on the partner.

Respondents were consequently deemed to have experienced coercive controlling violence if they reported both high levels of domestic violence and abuse experience, and that this had a significant impact upon their lives. Significant impact was measured using a combination of three separate scales that relied on the respondent to disclose his feelings about the abuse and whether he responded to it by changing his behavior or attitude toward his partner. As a result, of the 370 men who reported experiencing behaviors from the list, 31 were classified as experiencing a high rate of abuse with high impact, another 17 as experiencing a low rate of abuse with high impact, and another 142 as experiencing a low rate of abuse with low impact, with those 3 groups totalling 190 men reporting abuse that had an impact on their lives. Of these, only that first 31, or 6.13% of men reporting abuse, were classified as victims of abuse. 

The study also measured respondents’ disclosure of perpetration, but a different standard was used for that: All they had to do was admit engaging in up to 5 incidents of any combination of items on that same list of behaviors during their lifetime. The researchers did not ask the respondents whether these incidents were perpetrated against one partner or divided among multiple past partners. They did not gather information to measure the impact of these reported behaviors on the respondents’ partners’ lives. They also did not differentiate based on age. An elderly man would certainly have had many more years of opportunity to experience intimate partner relationships than a 20 year old man, but both would have been labeled perpetrators based on that 5-in-a-lifetime standard.

If a respondent reported refusing to do housework or having a lead foot while driving he would be as much labeled a perpetrator as one who reported having threatened his partner with physical harm, berated and insulted her, or beat her dog to cause her emotional distress. If a respondent described having attempted suicide to escape his partner’s abusive behavior and leaving a note to explain his reason, that would also be treated as an incident of perpetration.

Even with this massive double standard and failure to consider abuse as a potential contributing factor to addiction and suicidal ideation, the researchers were only able to label 38 men perpetrators - 7 more than they labeled victims based on their much stricter standards.

Less than half of the men labeled victims were on the list of men labeled perpetrators.

In their abstract, the researchers emphasized that group over all of their other data, while de-emphasizing the majority of men who had experienced coercive controlling behaviors, with the following statement:
“...only 4.4% of the men experienced coercive controlling violence and of those nearly half also reported perpetration against their partner.”

They stated in their conclusion that this should influence approaches to this phenomenon taken by clinicians and domestic violence advocacy agencies. You know, because if you ignore over 90% of men’s experiences of coercive, controlling behavior, then you can claim very few men are victims of it... and if you ignore the majority of the small group you acknowledge as victims who are innocent of any perpetration even under the loosest standards of measurement, you can say men who are victims of coercive control by women are just as bad as their perpetrators. This is the conclusion the layman is supposed to reach from research that shows that a significantly higher percentage of men than the researchers wanted to acknowledge experience this type of abuse, and more than half of those who experience the worst of it do so without lashing out at their partners in a similar fashion even as little as 5 times in their lives. 

How twisted is that?

Such academic chicanery is a tool for ideologues to create a body of false academic conclusions they can use to promote their political narratives. These are then fed to whichever media is most willing to disseminate them without questioning them. They’re published and publicized by that media, with many outlets putting out almost the same article, making the same observations about the same parts of the reports they’re endorsing.

The narratives are then picked up, carried, and defended by loyal followers of the ideology, using shaming language and blame against critics of the methods used to arrive at the report’s stated conclusions. They’re fed to students to influence their future professional work. They’re fed to legislators as the ideologues who promote them agitate for changes in law and policy. Then those ideologues, including those students when they graduate, benefit through initiatives that employ as administrators, advisers, and support staff people who can claim to be experts on the problems they’ve convinced those legislators needed to have money thrown at them… all at taxpayer expense, of course.

Become a patron to

Unlock 820 exclusive posts
Be part of the community
Listen anywhere
Connect via private message