LONDON, Ontario – One of the reasons the scientific method has been so successful over the centuries is its reliance upon the ability of researchers to reproduce outcomes as a way to confirm prior experimental results. Repeatability and confirmability are crucial elements of any good experiment, study or finding.

When it comes to the conventional wisdom that informs a lot of anti-pornography rhetoric, one 1989 study from Arizona State University frequently is cited to assert pornography viewing negatively affects men’s opinion of the attractiveness of their real-life partners.

In the 1989 study, evolutionary psychologist Douglas Kenrick had a group of college students examine “a photograph of a nude female after being exposed either to control stimuli (abstract art or other average nudes) or to photographs taken from popular erotic magazines.”

In the second experiment conducted as part of the study, “male and female subjects were exposed to opposite-sex erotica” and there was “an interaction of subject sex with stimulus condition upon sexual attraction ratings.”

Kenrick’s findings suggested there was a marked difference between the responses of male and female subjects, the chief difference being one that has dominated gender-bifurcated notions of porn’s differing impact on male and female viewers in the years since the study was published.

“Decremental effects of centerfold exposure were found only for male subjects exposed to female nudes,” the researchers claimed. “Males who found the Playboy-type centerfolds more pleasant rated themselves as less in love with their wives.”

For many people, these results resonate with their intuitive sense of what porn is: An objectification of women that causes men to be less interested in the “real” women with whom they interact daily.

Kenrick, the man leading the study, took the results to heart in his personal life.

“The lesson I took out of my own research is that I stopped looking at Playboy,” Kenrick said, “and I didn’t really miss it.”

In the years since Kenrick’s study was published, many other social scientists have produced similar results, although for the most part theirs have not been clinical studies, but meta-research based on survey data like the General Social Survey or the Portraits of American Life Study.

So, what happens when a researcher tries to replicate Kenrick’s study in a clinical setting? Well… Not much, quite frankly.

Researchers led by Lorne Campbell from the University of Western Ontario (UWO) were unable to replicate Kenrick’s results in multiple attempts. Naturally, as with any generally accepted experimental result that cannot be replicated in subsequent experimental duplications, the question is “Why can’t the results be reproduced?”

While a lot of observers seem convinced there’s some massive cultural or societal shift in attitudes toward porn to which the discrepancy in experiments can be attributed, the far more likely (and obvious, frankly) cause is one that hampers a great deal of research: The subject pool in the original study was too small and too narrow in demographic to be representative of the general population.

In Kenrick’s study, only 63 college students participated; in the UWO study, the subject pool was 10 times that size. While there are many differences between “studies” and “surveys,” the two do have something in common: The confidence in the result varies according to the sample size of the subject or respondent pool.

Regardless of the reason behind it, Kenrick said he’s “shocked” by the inability of UWO researchers to replicate his findings, which he said he trusts because he “analyzed this data myself.” Still, to his credit as a scientist and academic, Kenrick doesn’t disparage the UWO study, calling it “one of the cleaner, nicer nonreplications.”

This openness to the OWU outcome doesn’t mean Kenrick is about to concede his original study produced an unreliable result, of course, regardless the size of his subject pool.

“Maybe the damage has been done,” Kenrick said, speculating on why his study’s results could not be reproduced. “It used to be that looking at erotic magazines was a special event. People did it once a month if they had a subscription to Playboy.”

Really, Doug? Back in the day, men looked at Playboy the day it arrived in their mailbox, then stuck it in a growing pile of back issues where it collected dust until its day of recycling reckoning came along? Men in the late 1980s never watched porn on VHS? Never subscribed to more than one porn magazine?

Upon further review, maybe the problem wasn’t sample size, but the insanely ill-informed assumptions formed by the guy evaluating the data.

Speaking of sample size, has Kenricks ever discussed these things with even one man over the age of 45 who has been consuming porn throughout his adult life? If he were to do so, I suspect at least one of us would laugh uproariously right in his face about this notion that porn viewing was more “special” in the late ’80s, or the suggestion we had only one option (one softcore option, no less) for adult entertainment at a time when the market for porn on VHS had already been flourishing for several years.

For all its proven merits, the scientific method is ultimately only as reliable as those applying it. If a researcher analyzing data brings to the act a host of flawed assumptions, there’s just not a whole lot the raw data can do to make the wrongheaded researcher see the errors of his ways.