We all know a bullshitter. They can shoot off explanations and rationales for just about anything — even if they don’t know what they’re talking about. They’re not liars, purposefully hiding the truth, but they certainly don’t care if what they’re saying is true or not. Scientists have studied the phenomenon before, digging into how we perceive bullshit and its consequences. But what makes us actually bullshit in the first place? That’s what psychologist John Petrocelli explores in a new paper in the Journal of Experimental Psychology, aptly titled “Antecedents of Bullshitting.”

Let’s Talk About Jim

The work consists of two separate experiments. In the first, Petrocelli wanted to dive into how bullshitting behavior is impacted by three things: someone’s knowledge of whatever they’re talking about; the social obligation they feel to provide an opinion on the topic; and what he calls the ease of passing bullshit — how our audience’s knowledge of the topic plays into how likely they are to call bullshit.

So he used Amazon Mechanical Turk, a crowdsourcing job-task website, and recruited nearly 600 participants. These people were split up into groups that focused on those three different variables. However, they went in assuming the study was testing our opinions on what motivates the behavior of others. Specifically, they were told about a man, Jim, who’d been running for a seat on his local city council. Though he was leading in some polls, he’d decided to pull out of the race a month before the election. Participants were then asked to think about why Jim might’ve taken himself out of the running and given the chance to list five reasons why they thought Jim dropped out.

Here’s where things get interesting. Remember those three variables we talked about earlier? After getting these initial instructions, certain groups then got extra instructions that aligned with those variables. To test how social obligation plays into bullshitting, Petrocelli specifically told some participants they didn’t need to share their opinions if they didn’t want to. To look at how ease of passing bullshit impacts things, another portion of the participants were told that the people looking at their answers, the coders, knew Jim extremely well. And to study how someone’s knowledge of a topic affects their decision to bullshit, before they gave their answers, some of the participants had read a series of statements about Jim, gleaned from a personality test he had taken.

Things went kind of like you’d expect. People were way less likely to bullshit when: they got the low-down on Jim beforehand; there was less pressure to provide an opinion on Jim’s decision; and when people thought the coders knew Jim really well. There was even an interaction with social obligation and audience knowledge. People who felt more pressure to give their thoughts on Jim were more likely to bullshit, even if the ease of passing bullshit was low — i.e., if they were told their coders knew Jim well. But if the pressure was off? No bullshit necessary.

I Call Bullshit

To dig deeper into the bullshit phenomenon, Petrocelli carried out a second experiment. This time, he invited participants in four groups to give their opinions on controversial topics like affirmative action quotas and capital punishment. He then told three of these groups they’d have to discuss their thoughts with someone else — the fourth was a control that didn’t have a discussion buddy. One group was told they’d be chatting with someone who had opinions similar to them. A second group was told they’d discuss with someone who had an opposing view, and the last group didn’t know what the other person thought. Petrocelli hypothesized people in the second and last groups would be less likely to bullshit.

Again, it went as you might expect. When you’re up against someone who might call you out on your bullshit — as in the second and third groups — you’re far less likely to spew it.

Though it’s not the most robust study — some limitations include not accounting for the cognitive abilities of participants and also the fact that these “discussions” were pretty passive — it’s an interesting springboard for future work into what makes us okay with bullshitting.