As long-time World of Psychology readers already know, a researcher has a lot of latitude in how they design a study to “encourage” a predetermined outcome. Researchers generally don’t recognize this as an inherent bias problem, because virtually all researchers do it to one degree or another (or have done it at one time or another in their career).
The relationship between mental illness and violence is one area of contention among researchers, with most research showing only the smallest of correlations between the two. The real risk factor for violence remains — and has always been — substance abuse, not mental illness.
Recently it was suggested that those with bipolar disorder are at greater risk for committing violence. So we took a look at some of the research to see how good the studies are that suggest such a connection.
How You Define Things Helps Pre-Determine Your Results
Researchers can get halfway to their desired outcome by setting up the definition of their terms in ways that most benefit their hypotheses. I always check this first, because it’s so easy for a researcher to manipulate this variable without raising the eyebrows of anyone except those most interested in digging into the data (like I like to do).
Take, for instance, the definition of “violent crime.” You’re welcomed to use any definition you want. But if you’re going to study a controversial relationship of this nature, you’d think you would use a well-known, accepted definition to ensure the most objective, generalizable results. You know, like using the U.S. Federal Bureau of Investigation’s definition of violent crime:
In the FBI’s Uniform Crime Reporting (UCR) Program, violent crime is composed of four offenses: murder and nonnegligent manslaughter, forcible rape, robbery, and aggravated assault.
That’s it, just those four things. But being from Sweden, here’s how Fazel et al. (2010) define it:
In keeping with other studies, violent crime was defined as homicide, assault, robbery, arson, any sexual offense (rape, sexual coercion, child molestation, indecent exposure, or sexual harassment), illegal threats, or intimidation. Attempted and aggravated forms of offenses, where applicable, were also included.
Those other studies? Also conducted by the same first-author (Fazel & Grann, 2006; Fazel et al., 2009), with neither study providing any justification for this extensive list of crimes — some of which may not even be perpetrated against an individual (e.g., arson). ((In fact, the second study cited also cites the 2006 study, which provided no rationalization for the crimes included. As an aside, I find it a bit disingenuous when an author cites their own work for justification of a definition. It’s very circular reasoning, especially when those earlier studies actually don’t provide any further clarity on why those specific crimes were chosen.)
When I reached out to the Swedish National Council for Crime Prevention for clarification, a spokesperson noted that Sweden doesn’t have an official definition of “violent crime” like the U.S. does. Instead, they have a much broader category called “Crimes Against the Person,” which includes not only violent crimes, but also non-violent offenses (such as defamation and “intrusive photography”).
By including a broader definition of “violent crime” than most, researchers for this study ensured they would lasso more people who’ve been convicted of these additional crimes. And while it may be interesting to note if someone with a mental illness is more likely to commit a crime against property (versus a person), that’s a very different research question than looking into one where we are most concern about a person’s inclination to commit a “violent crime.”
What Did They Find, Even With This Broad Definition?
In this study with the broader definition of “violent crime,” the researchers still didn’t find much of an increased risk for people with bipolar disorder to commit a violent crime — unless they were also abusing a substance (like drugs or alcohol):
Instead, the association between bipolar disorder and violent crime seemed to be largely mediated by substance abuse comorbidity. The risk increase in patients with bipolar disorder and substance abuse comorbidity was more than that found in a related study of schizophrenia.
At least for bipolar disorder, the authors of this study say, “[…] we found no increased violence risk in patients having bipolar disorder without substance abuse comorbidity […]” In other words, once you remove substance abuse from the equation, people with bipolar disorder appear to be at little greater risk in committing acts of violence than someone in the general population.
The risk associated with a bipolar diagnosis per se appears low; it was minimal compared with that in general population controls when there was no comorbid substance abuse, and there was no association when violence risk in patients was compared with that in unaffected siblings.
The day we stop blaming violence on mental illness and repeating this simplistic falsehood is the day we can move on to real solutions to the violence problem in America.
Fazel et al. (2010). Bipolar Disorder and Violent Crime New Evidence From Population-Based Longitudinal Studies and Systematic Review. Archives of General Psychiatry, 67, 931-938.
Fazel S. & Grann M. (2006). The population impact of severe mental illness on violent crime. Am J Psychiatry, 163, 1397-1403.
Fazel S, La°ngström N, Hjern A, Grann M, Lichtenstein P. (2009). Schizophrenia, substance abuse, and violent crime. JAMA, 301, 2016-2023.