It’s safe to say that porn watching is widespread these days, but what we can’t say with certainty is how all that porn is changing the way we behave. Scientists still aren’t sure: Some suspect that porn seeds ideas into our heads about how sexual relationships are supposed to be, but few have been able to show this through research. Now, however, a study on a very niche type of porn in the journal Psychology of Popular Media Culture shows this phenomenon at work.

The study, led by James Cook University psychologist Dan J. Miller, Ph.D., investigated this idea by exposing half of the study’s 418 male participants to a very specific type of porn known as “taxi porn” — essentially, clips in which a male taxi driver has characteristically “porn-like” sex (rough and unprotected) with a female stranger looking for a ride.

He hypothesized that men exposed to this type of porn might be more inclined to think that women are into that kind of sex, an idea steeped in “sexual scripts theory.” This theory, one of the leading explanations for how porn changes our sexual behavior, suggests that porn provides us with “scripts” that we then play out in real life.

taxi cab
In taxi porn, a male driver usually propositions and has sex with a female stranger looking for a ride.

The reason the team used “taxi porn” is because they wanted to account for the issue of “past exposure,” so that they could know they were actually measuring the effect of when porn was viewed. Similar recent studies have been invalidated by failing to take this factor into account. If a scientist wants to know how watching porn changes a person, they must account for the porn being shown in the experiment as well as porn that person has watched before; in this case, the team assumed that taxi porn was niche enough that not too many men would have watched it recently (which, they later found out, was incorrect — about one-third had done so). As part of the study, the participants revealed the types of porn they’d viewed in the past six months.

While the control group watched educational videos from YouTube, the experimental group watched one of two 11-minute “taxi porn” clips from YouPorn: either “Fake Taxi, Stranded French Tourist Earns Extra Cash,” or “Fake Taxi, Heavy Metal Grupie [sic] Likes it Hard and Rough.” In both clips, a young woman has rough, unprotected sex with her driver and appears to like it. After viewing, all the participants were asked what they thought could happen in two written hypothetical scenarios. The first mirrored the premise of the porn scenes, in which a taxi driver propositions a female passenger; in the second, a male boss hits on his female employee.

Sure enough, there was a link between past porn exposure and the likelihood that a participant would think the woman in either of those hypothetical scenarios would engage in “porn-like” sex with the taxi driver or the workplace boss — but it wasn’t rooted in the videos the participants had just watched. Rather, the scientists saw a link between people who said they’d viewed taxi-themed or workplace-themed porn in the past six months.

male boss
Men who had seen workplace-themed porn were more likely to think a woman would engage in porn-like sex with her male boss.

The results didn’t allow the scientists to draw any firm conclusions about the mechanism by which porn affects male perceptions about women, but they did present evidence supporting the idea that porn changes male perceptions in a broad way. It’s possible, they hypothesize, that the reason only men who had watched past taxi- or workplace-themed porn had differing views was because of the “sleeper effect” — the idea that it takes a while for ideas to take root after being exposed to them. It’s also a possibility that men only select porn that’s already in line with their existing beliefs.

“The study provides some evidence that pornography can influence consumers’ judgements of social reality, by affecting consumers’ perceptions of the likelihood of women enthusiastically engaging in the kinds of sexual practices commonly depicted in pornography,” the team writes.

If this is actually the case, then our porn use is becoming increasingly problematic. In a worrying illustration of this study’s findings, a UK man admitted in June to pretending to be a cab driver so he could rape women after being inspired by taxi porn.

A lot of the porn out there is geared toward heterosexual men, and, as the team points out, a lot of that porn shows behavior that not only diminishes the importance of consent but also is straight-up dangerous, psychologically and medically (there’s a reason condoms are such an important part of sex ed). Of course, we’re never going to get rid of porn, but the hope is that, as more studies like this show the effect of porn on viewers, we might get better at shaping porn to show sex that doesn’t just titillate but educates as well.