Scary Bridge Study: Can fear make you horny?

[Regina] (0:00 - 0:10)
I went out with a clinical psychologist, and for our second date, he gave me the following options, rollerskating, go-karting, or axe-throwing.

[Kristin] (0:10 - 0:46)
I think he might have known about the scary bridge study then. But Regina, axe-throwing, what kind of date is that?

Welcome to Normal Curves.

This is a podcast for anyone who wants to learn about scientific studies and the statistics behind them. It's like a journal club, except we pick topics that are fun, relevant, and sometimes a little spicy. We evaluate the evidence, and we also give you the tools that you need to evaluate scientific studies on your own.

I'm Kristin Sainani. I'm a professor at Stanford University.

[Regina] (0:46 - 0:52)
And I'm Regina Nuzzo. I'm a professor at Gallaudet University and part-time lecturer at Stanford.

[Kristin] (0:52 - 0:57)
We are not medical doctors. We are PhDs, so nothing in this podcast should be construed as medical advice.

[Regina] (0:57 - 1:14)
Also, this podcast is separate from our day jobs at Stanford and Gallaudet University.

Kristin, today we're time-traveling back to 1974. Ooh.

Mm-hmm. So, let's go put on some bell-bottoms, huh, and get ready to get groovy.

[Kristin] (1:14 - 1:17)
You know, Regina, I'm more of a miniskirt kind of person myself.

[Regina] (1:19 - 1:26)
Okay. On you, that would work, actually. Sequined disco miniskirt, yep.

[Kristin] (1:26 - 1:27)
Totally age-appropriate, Regina.

[Regina] (1:28 - 1:52)
Well, Kristin, you've also got to wear it with me on the tall, scary bridge. Part of the deal.

[Kristin]
Okay. And why?

[Regina]
Uh-huh. Because the paper we're looking at today, classic psychology study, sometimes known as the bridge study or the suspension bridge study or the scary bridge study or, as I like to think of it, the scary sexy bridge study.

[Kristin] (1:53 - 1:59)
I've never heard of that study before, Regina, but it sounds just like the kind of study that we love to talk about on this podcast.

[Regina] (2:00 - 2:11)
Right? Doesn't it? It's essentially a story, Kristin, about whether being scared makes other people magically more attractive.

Basically, it's, does fear make you horny?

[Kristin] (2:11 - 2:24)
Wait, so Halloween parties, possibly not just for a dress-up. They're a prelude to hooking up. And Halloween is coming up, Regina, so I feel like this episode is part public service announcement.

[Regina] (2:25 - 2:28)
We've got perfect timing, actually, don't we?

[Kristin] (2:28 - 2:28)
Yeah.

[Regina] (2:28 - 3:42)
Now, Kristin, you said you haven't heard of this study, so do you mind if I read how one article describes the gist of it? Because I think it'll whet your appetite.

[Kristin]
Okay, go for it.

[Regina]
A man named Meyer stands at one end of a 450-foot suspension bridge located in a British Columbian park. The bridge is made of wooden planks attached to wire cables and spans a deep canyon. Like many who visit the park, Meyer decides to test his nerve by crossing the narrow bridge.

When he reaches the middle, a gust of wind causes the bridge to sway, and he glances down at jagged rocks and swirling water 200 feet below. When Meyer looks up again, he notices an attractive woman. Of course he does.

Under different circumstances, Meyer might have been only moderately attracted to her, but in his aroused state, he is strongly attracted to her. He introduces himself and later asks for her phone number. They date, fall in love, and eventually marry.

The end.

[Kristin] (3:43 - 3:57)
All right, Regina, first of all, that bridge sounds super scary, and I guess that's one way to meet men?

[Regina]
One way, yeah.

[Kristin]
But is this evidence-based, Regina?

Because before I go and try this out, I need to know, does this actually work?

[Regina] (3:58 - 4:13)
Uh-huh. Well, that is exactly what we're going to look at today, and the claim that we'll examine is that experiencing heart-pounding fear makes men see women as more attractive than they would be otherwise.

[Kristin] (4:14 - 4:25)
So the theory is that being afraid and being horny get mixed up in your brain somehow?

[Regina]
Exactly. That's exactly it.

[Kristin]
Does this apply only to men or also to women?

[Regina] (4:26 - 5:02)
Well, in this paper, they only studied men looking at women, maybe because it was 1974 and no one was gay, and women were just sex objects standing around on bridges, no desires of their own. Yeah. But a better part of the 1970s vibe, in addition to the miniskirts, are the classic old-fashioned stats methods, and those are what we're going to cover today, t-tests, chi-square tests, and also our favorite, statistical sleuthing.

[Kristin] (5:03 - 5:22)
Hmm. So this sounds like Statistics 101, but better, with horny men in Halloween costumes. I love it.

[Regina]
Kristin, I like the way you think. I do believe you're getting better at this.

[Kristin]
I know.

I think I'm getting better at the sex references. I think I'm embracing our come-for-the-sex, stay-for-the-stats philosophy.

[Regina] (5:23 - 5:28)
Kristin, I cannot tell you how honestly proud I am to hear you say that.

[Kristin] (5:29 - 5:33)
It's all your influence, Regina.

All right, but back to the study. Tell us about the study. Mm-hmm.

[Regina] (5:34 - 5:43)
It's a classic, often taught in psychology class, and I personally kind of love the pop psychology angle of this one.

[Kristin] (5:44 - 5:46)
Oh, have you tried this trick out in real life, Regina?

[Regina] (5:46 - 6:00)
Not the bridge, but I did do another version. I went out with a clinical psychologist, and for our second date, he gave me the following options. Roller skating, go-karting, or axe-throwing.

[Kristin] (6:00 - 6:07)
I think he might have known about the scary bridge study, then. But Regina, axe-throwing. What kind of date is that?

[Regina] (6:08 - 6:14)
Yep, exactly. And that is actually the one that I chose, which made for an interesting date.

[Kristin] (6:14 - 6:24)
How have I never heard this story before? Okay, so I hope this wasn't like axe-throwing at people, like when you try not to hit the woman. Like, what were you throwing the axe at?

[Regina] (6:24 - 6:26)
A bullseye at a bar.

[Kristin] (6:26 - 6:34)
Oh, that sounds really safe. So we've got drunk people throwing axes. Great combination.

Brilliant.

[Regina] (6:35 - 6:38)
Well, I did not personally lop off anyone's toes.

[Kristin] (6:38 - 6:39)
Good.

[Regina] (6:39 - 7:04)
Mm-hmm. And I actually won. I've got a photo of me getting the bullseye, actually.

[Kristin]
Wow!

[Regina]
Uh-huh. I'll share it with you.

Yeah, the whole thing was good.

[Kristin]
I think we should post the photo on the show notes, Regina. I'm so impressed.

Maybe if we make it really, really small. But what about you, Kristin? Do you mix fear and dating?

[Kristin] (7:04 - 7:10)
Does motorcycle riding count? That definitely counts, yeah.

I once rode 13 hours on the back of a motorcycle to Crater Lake in Oregon, and then back again.

[Regina] (7:10 - 7:14)
Oh, the bad boy motorcycle grad school boyfriend, huh?

[Kristin] (7:15 - 7:22)
Exactly. But, you know, Regina, it just made me scared. Now, I did cling to him for dear life, so maybe it made him horny?

[Regina] (7:22 - 7:24)
Oh, I think it definitely made him horny.

[Kristin] (7:25 - 7:31)
Regina, why would fear make you horny? Like, why would other people look more attractive if you think you were about to die?

[Regina] (7:32 - 7:49)
Excellent question. You know, originally, I thought this might have been evolution's weird way of making sure that men pass on their genes before they get mauled by the saber-toothed tiger. Like one last-ditch attempt to keep the whole species alive before they're dead.

[Kristin] (7:50 - 8:03)
Oh, that's an interesting theory, and that's going to be some interesting sex right there. So is this like when the tiger is a mile away and stalking you, or literally while you're being mauled at the same time, and if the latter, then how does that work?

[Regina] (8:04 - 8:14)
You know, I confess I had not really worked out the mechanics of it until you asked just now, but I'm sure there's a Reddit group for it.

[Kristin] (8:15 - 8:24)
We're talking evolution cavemen, Regina, not alternative lifestyle. But, Regina, what does the science say? Is that actually the evolutionary reason?

[Regina] (8:25 - 8:34)
Sadly, no. Psychologists have a more subtle and more boring hypothesis. They call it misattribution of arousal.

[Kristin] (8:35 - 8:36)
Oh, that sounds jargony.

[Regina] (8:37 - 9:10)
Doesn't it? But the idea is pretty simple. You're on the scary bridge, and you might plunge to your death.

So you're jittery, your heart's pounding, you've got butterflies in your stomach, and do you know what else gives you butterflies and a pounding heart? Falling in love. Or wanting to have sex.

[Kristin]
Oh, interesting.

[Regina]
So the idea is that you notice your physiological arousal from fear, but you misattribute it to being horny for the potential date next to you. That's kind of it, nutshell.

[Kristin] (9:10 - 9:20)
Psychology comes up with weird theories. I'm going to need evidence. And the bridge you described earlier, that's the bridge they used in the study, the shaky suspension bridge?

[Regina] (9:21 - 9:36)
450-foot suspension bridge over a deep canyon, narrow, low handrails, swaying with every step. Beautiful, right? It's called the Capilano Suspension Bridge.

It's in a park in Vancouver, Canada.

[Kristin] (9:36 - 9:46)
Yeah, I think I would be excluded from this study, because I can tell you right now I am not going to be walking over that bridge. My fear of heights comes out when railings are low and things sway.

[Regina] (9:46 - 10:05)
That is actually a good survival instinct. I've got a bit of fear of heights myself. But Kristin, apparently many people do not agree with us about this whole swaying bridge thing, because Wikipedia says the Capilano Bridge gets 1.2 million visitors a year.

[Kristin] (10:05 - 10:20)
Wow, that is a lot. But let's talk statistics, Regina. That's actually the only way I could possibly get on this bridge is with statistics.

So what is the actual risk of falling from the bridge? How many people have plunged to their deaths over the years?

[Regina] (10:20 - 10:29)
I thought you would ask that. I looked hard, but I could not find any reports of people falling from the bridge itself. Oh, you're kidding.

[Kristin] (10:30 - 10:38)
So it's actually really safe, even though it seems scary, and I might be able to get on it. It's statistically, like, safer than taking a bath.

[Regina] (10:40 - 10:44)
Which is also sexy and great for dates.

[Kristin] (10:45 - 10:47)
Like third date material.

[Regina] (10:47 - 10:48)
Yeah, yeah, at least.

[Kristin] (10:49 - 11:04)
After the axe throwing. After the axe throwing. Okay, so there was a paper involved, really.

This is a podcast about research papers and not just boys and daredevil stories or daredevil boys stories. So, Regina, let's talk about the paper.

[Regina] (11:04 - 11:28)
Oh, yes, the paper. Back to reality. All right, the two authors were Donald Dutton, a psych professor at University of British Columbia, and Arthur Aron, his postdoc.

And it was published in Journal of Personality and Social Psychology in 1974. They reported on three experiments, actually, but the first one is the famous bridge study.

[Kristin] (11:28 - 11:46)
That's the one with the swinging bridge. And, Regina, 1974, this is a classic paper. I want to point out that we swing wildly on normal curves between papers just published, like breaking news, and classic papers published half a century ago.

We mix it up. We keep it interesting.

[Regina] (11:47 - 12:17)
I personally love the old papers, Kristin, because there's something so, I don't know, charming and nostalgic about them. Back when stats was done by hand and women were sex objects. But some of these studies, actually, they've got ballsy designs.

I mean, a swinging bridge over a 200-foot canyon, it's amazing they pulled this off. Because what if someone just stumbled a bit during the study, right? That would not be good.

[Kristin] (12:17 - 12:24)
That would not be good. And I'm guessing that this study was done before IRBs, institutional review boards, ethics committees, perhaps?

[Regina] (12:24 - 12:55)
Uh-huh. Bingo. IRBs did not really start until the late 1970s.

I did check that out because I was curious myself. And, you know, it's kind of interesting, though, to look at this study with our modern eyes. And now we would ask, are you exposing participants to harm by putting them on this crazy swinging bridge over a canyon?

Do they actually give proper consent? And should you even be doing this kind of study? It's the sort of thing that ethicists like to talk about now.

[Kristin] (12:56 - 12:59)
The study could never be repeated, clearly, Regina.

[Regina] (12:59 - 13:05)
Uh-huh. Well, half of the study was safe because they did have a control bridge that was much more tame.

[Kristin] (13:05 - 13:09)
Oh, good. We love controls. And what was the control bridge?

[Regina] (13:09 - 13:18)
It was a super safe one just up the river, wide, solid cedar, 10 feet above a ditch with a little rivulet, completely boring.

[Kristin] (13:19 - 13:21)
So I could run over this bridge without a moment's thought.

[Regina] (13:22 - 13:40)
Absolutely. But in the middle of both bridges, however, was the experimental stimulus that you would run into, which was an attractive young female interviewer. The same pretty woman stationed at different times on both the scary bridge and the safe bridge.

[Kristin] (13:41 - 13:43)
Do we know anything about her?

[Regina] (13:43 - 14:00)
No, no picture, no information. And, you know, they don't say how she was dressed, but since it's 1974, I'm picturing miniskirt and white plastic knee-high go-go boots. But that's just my head canon.

They didn't say.

[Kristin] (14:01 - 14:08)
You know, she had to be pretty brave to stand out there for so long and also deal with a bunch of random men hitting on her.

[Regina] (14:08 - 14:27)
Uh-huh. They said she was blinded to the study, so I don't think she knew it was about horniness. And I'm not sure if it's better or worse for her, actually, not knowing.

[Kristin]
Regina, did she at least get to be a co-author on this study?

[Regina]
Nope. Anonymous and no authorship.

[Kristin] (14:28 - 14:33)
I feel like she should have gotten credit for this because people have done a lot less for authorship.

[Regina] (14:33 - 14:34)
For sure.

[Kristin] (14:35 - 14:38)
Now, Regina, did they randomize men to the two bridges?

[Regina] (14:38 - 14:47)
You would think so, but no. They simply used the guys who already happened to be walking across each bridge.

[Kristin] (14:47 - 15:02)
Oh, that's kind of a bummer. So, it was more of a convenience sample. And, well, I guess I can see logistically it would be hard to pull off randomization, right?

You would, what, randomize them in the lab and then drive them to the different bridges? I get that it's logistically complicated.

[Regina] (15:03 - 15:22)
Definitely a convenience sample because they also did not have much in the way of inclusion and exclusion criteria because the pretty interviewer stopped every guy who looked like he was between 18 and 35 years old and who was not already walking with a woman. That's it.

[Kristin] (15:23 - 15:43)
Oh, okay. So, not randomized, but that means we have very different samples because men who are choosing to walk across a terrifying suspension bridge are likely quite different from the men on the boring control bridge. Like, maybe they are more risk-takers or just into danger, right?

So, not a perfect comparison.

[Regina] (15:44 - 16:05)
The investigators themselves pointed that out, actually, Kristin. They said they might have been thrill-seekers, but the investigators also pointed out that the scary bridge might have more tourists and that tourists might be more sexually deprived or that the tourists might not have known they were sexually deprived until they saw a pretty woman.

[Kristin] (16:07 - 16:10)
That is a rather bizarre and convoluted theory, Regina. Okay.

[Regina] (16:12 - 16:18)
I feel like, Kristin, you've identified problem number one with the study, and by the way, there will be many problems to come.

[Kristin] (16:19 - 16:26)
Right. So, problem number one is the bridge is confounded with the participants' personality and risk-taking behavior.

[Regina] (16:27 - 16:31)
Let's call it the adventure dude confounder problem.

[Kristin] (16:31 - 16:32)
I like it.

[Regina] (16:33 - 16:51)
Now, the investigators tried to fix this adventure dude confounder problem in Experiment 2 by using only the scary bridge. They got guys either on the bridge or well after they had crossed the bridge and had cooled down, but we'll talk about all of that later.

[Kristin] (16:52 - 17:00)
Oh, cool. Great that they did that.

But, Regina, what did they actually do on the bridges? Was there flirting, kissing, sex? Like, what happened?

[Regina] (17:00 - 17:15)
Not quite. The pretty interviewer would approach the potential participant and say she was doing a project for her psych class on, quote, the effects of exposure to scenic attractions on creative expression.

[Kristin] (17:16 - 17:25)
Oh, kind of hiding the plot there.

[Regina]
Uh-huh.

[Kristin]
And what did the men actually have to do, and how did they measure if the men were hot for the women then?

[Regina] (17:26 - 18:04)
Uh-huh. Yeah. Okay, they used two outcomes.

The first was through a questionnaire, but this was no ordinary questionnaire.

[Kristin]
Oh, do tell.

[Regina]
Uh-huh.

On the second page was a black-and-white drawing of a woman standing in a doorway. I'll put the image in the show notes, but I'll describe it here. One hand was covering her face, and the other was gripping the side of the door.

And it's kind of ambiguous. You can read into it whatever you want. And the men were asked to write a, quote, brief dramatic story about what they imagined was going on in the image.

[Kristin] (18:04 - 18:20)
Oh, wow. So they had to write a creative writing story on a wobbly bridge with low handrails? I'm shocked that anyone agreed to do this study.

I'm thinking she was really pretty. And I want to read these stories. Did anyone ever publish the stories?

[Regina] (18:21 - 18:33)
I have not found them published anywhere, but I will contact one of the authors to see if he still has them lying around somewhere. Sure, 50 years later.

[Kristin] (18:33 - 18:41)
Yeah, 50 years. Okay, maybe not.

Although, admittedly, if it was my office, I could probably dig them out of a folder, like hid it in a box under another box somewhere.

[Regina] (18:42 - 19:03)
I've seen your office. I believe it. So that black-and-white drawing, Kristin, is actually part of a standard psychological test called the thematic apperception test.

And it's supposed to get at people's hidden motives and drives the things that psychologists love, fear, desire, aggression.

[Kristin] (19:04 - 19:07)
Oh, like an inkblot test, but with a full written story.

[Regina] (19:08 - 19:21)
Right, exactly. And the investigators wanted to see whether the men would write sexy stories about this drawing. And believe me, this little drawing was not explicitly sexy at all.

[Kristin] (19:21 - 19:28)
I don't know, gripping the doorway. But how do you measure whether the essay had sexy content?

[Regina] (19:29 - 20:14)
Good question. And they described it in the paper. They said that they gave each story a score of one to five based on how much sexual content it had.

And this was their scale. They described that any mention of sexual intercourse at all meant the story automatically got a score of five points. Makes sense.

Sexual references, however, that were a little less explicit and a little bit more hidden were given lower scores. So a story that used the word lover would get a score of four. And one that used kiss would get a three.

One with the word girlfriend would get a two. And a story with no sexual content at all was assigned a one.

[Kristin] (20:14 - 20:23)
Hey, Regina, this is kind of like our smooch scale, but more interesting. And I think just for this episode, we should switch to this scale when rating the claim for today.

[Regina] (20:23 - 20:25)
Ooh, let's do it. I like it.

[Kristin] (20:26 - 20:27)
And who rated these essays?

[Regina] (20:28 - 20:50)
Two skilled raters. They didn't say who. I assume it was the authors themselves.

Yeah, they did say that the raters were blinded to which bridge the men were on and that the inter-rater reliability was .87.

[Kristin] (20:51 - 20:56)
Oh, so for listeners who aren't familiar with inter-rater reliability, that score just tells us that they actually agreed a lot.

But to me, it still seems pretty subjective. So maybe we call this the subjective sexy scale problem.

[Regina] (20:57 - 20:58)
I like it.

[Kristin] (20:58 - 21:00)
And, Regina, what was the second outcome?

[Regina] (21:00 - 21:20)
Right. So after the questionnaire, the pretty interviewer thanked the guy and offered to explain the experiment in more detail later when she had more time. And she tore the corner off a sheet of paper, wrote down her name and phone number, and invited him to call if he wanted to talk more.

[Kristin] (21:21 - 21:21)
Ooh.

[Regina] (21:21 - 21:29)
And the investigators tracked whether the men took her number and then whether they actually called her later. Right.

[Kristin] (21:29 - 21:32)
They called her later to talk about this study.

[Regina] (21:35 - 21:51)
I know. You know, I'm torn on this one. I do think maybe they don't give men enough credit because not all men are horndogs, Kristin.

Maybe some called her because they actually wanted to talk about science. Maybe.

[Kristin] (21:52 - 21:56)
Have you met men, Regina?

I think this is your hopeful picture of men, but I don't think it matches reality.

[Regina] (21:56 - 22:13)
Oh, okay. Good point, good point. But there is another possible issue, and that is the cool factor.

Because maybe guys just thought a study on a scary bridge sounded cooler and wanted to learn more.

[Kristin] (22:13 - 22:18)
Oh, I could see that. So maybe legitimately on the scary bridge they were actually curious and called her back to talk about the science. Maybe.

[Regina] (22:19 - 22:32)
Uh-huh. Now the researchers actually tried to check out this cool factor thing. You know, is the scary bridge cooler than the boring bridge?

And what they did is run the exact same setup with a male interviewer.

[Kristin] (22:33 - 22:55)
Oh, this is great. So this was their attempt at what we would call a negative control. If the guys on the scary bridge also called the male interviewer back, more than guys on the safe bridge, that would mean it wasn't all about attraction.

It might be genuine curiosity about what the heck are you doing a study about on the middle of a swaying bridge. I can see that.

[Regina] (22:55 - 23:06)
Right, right. Unfortunately, Kristin, their numbers were so tiny. We're talking two guys on the scary bridge called back versus one guy on the safe bridge.

So we can't really tell.

[Kristin] (23:07 - 23:21)
Yeah, with numbers that small, you can't really draw any conclusions. But I do really want to give the researchers a lot of credit for thinking about this ahead of time, about this cool factor problem, and trying to put in a check. That was actually really smart of them.

[Regina] (23:22 - 23:25)
It was smart. I agree.

OK, I think we're ready now to talk about the results.

[Kristin] (23:25 - 23:59)
I can't wait, but let's take a short break first.

Welcome back to Normal Curves. Today, we're taking a closer look at the scary bridge study to find out whether fear makes you horny.

Regina, you were about to tell us the results. So what did they find?

[Regina] (24:00 - 24:11)
First of all, in the scary bridge, they approached 33 men and got 23 to fill out the questionnaire. And on the safe control bridge, they got 22 out of 33 to fill it out.

[Kristin] (24:12 - 24:21)
Wow. She must have been really pretty because actually that's quite a good response rate. I will point out, though, also it's a small N.

It's a small sample size.

[Regina] (24:21 - 24:41)
Yeah, exactly. But Kristin, the simple numbers made it easier for me to do some statistical sleuthing. I did it.

I was a little surprised.

[Kristin]
Uh-oh.

[Regina]
So to make this more fun, I'm going to give you the results.

And I want you to see if you can spot the arithmetic problem.

[Kristin] (24:41 - 24:44)
Ooh, statistical sleuthing quiz. I'm in, Regina.

[Regina] (24:44 - 25:15)
OK, let's start with the sexy story scores for the female interviewer. So as far as I can tell, each story got a whole number score from 1 to 5, where 1 meant no sexual imagery and 5 meant explicit sexual imagery. Now, they said they got 20 usable stories from men on the scary bridge, and those stories had an average score or mean score of 2.47. And from the safe control bridge, they got 18 stories with an average score of 1.41. You see the problem, Kristin?

[Kristin] (25:15 - 26:21)
OK, Regina, this actually involves a little mental math. And so I'm going to start with the mean of the scary bridge because you said the sample size was 20, and I love round numbers when we're doing math. OK, all right.

I think I see the problem, Regina. That mean of 2.47 is actually impossible. And let me walk everyone through why.

We get an average by adding up all the men's scores on this whole number scale and then dividing by 20. But that sum has to be a whole number, too. And when you divide a whole number by 20, only certain decimals are possible.

The result has to end in either .00 or .05 because 1 out of 20 is .05. It's like a nickel, right? There are 20 nickels to a dollar, so the result has to be an increment of .05. That means you can get a mean of 2.45 and you can get 2.50, but you actually cannot get 2.47.

[Regina] (26:22 - 27:03)
Right, no pennies, only nickels.

You've got such an interesting brain, Kristin, because you like division. I think in terms of multiplying better. OK, so another way to see this is to multiply the average by the sample size.

And if the math works out right, this is the total of all of the scores, which should be a whole number. So 2.47 times 20. And let me show you my little mental arithmetic trick for doing that.

Multiply by 10 first. So you just move the decimal one place to the right and you get 24.7. And then you double that, which is easy, so you get 49.4.

[Kristin] (27:03 - 27:04)
I love that mental math trick, Regina. Very good.

[Regina] (27:04 - 27:23)
Thank you, but if you add up a bunch of whole numbers, there's no way to get 49.4. So you know it's an impossible average. Same thing with the other group. 18 men, mean score of 1.41. 18 times 1.41 is 25.38, also not a whole number.

[Kristin] (27:23 - 27:26)
I'd probably have to put that one in Google, Regina, a little harder to do in your head.

[Regina] (27:27 - 27:33)
I did not do that one in my head, I confess. OK, so that one is also a weird average.

[Kristin] (27:33 - 27:51)
Regina, I love this trick because this isn't something really fancy. It's basic arithmetic. And a little detour for listeners, this kind of math check has a name.

It's called the GRIM test, short for granularity-related inconsistency of means.

[Regina] (27:51 - 27:54)
They did work hard to get that acronym.

[Kristin] (27:54 - 28:14)
They did, and points for the dark humor there, Regina. The GRIM test checks whether a reported mean is possible, given the sample size and the rounding rules, and it works for variables that are measured on an integer scale like this sexy essay scale. And there's even a website and an app where you can just punch the numbers in without having to do the math yourself.

[Regina] (28:14 - 28:53)
The GRIM test was actually invented by a couple of friends of ours, James Heathers and Nick Brown. James Heathers has a very nice podcast called Everything Hertz. James and Nick used the GRIM test to expose a ton of exactly these kinds of arithmetic problems in studies from a researcher named Brian Wansink, who studied nutrition at Cornell University.

And the whole thing is a really interesting saga. It's commonly known as Pizzagate. And, Kristin, I think we really should do an episode on Pizzagate.

[Kristin] (28:53 - 29:05)
We should. That is a great statistical sleuthing saga. And, Regina, Pizzagate, by the way, not related to Hillary Clinton conspiracy theories.

That's a totally different Pizzagate.

[Regina] (29:05 - 29:30)
No politics here. Absolutely not.

Okay, so back to our own statistical sleuthing. Kristin, I was trying to come up with some explanation for how all of this might have happened. And I realized that although the author didn't say it in the paper, they might have given each essay the average of two Rater scores.

So then each essay would not need to be a whole number.

[Kristin] (29:31 - 29:45)
Right, because there were actually two Raters. And if you average them, then each essay score could be a number between one and five in increments of 0.5. So you could get a sum of 49.5, for example.

[Regina] (29:46 - 30:28)
Exactly. Now that math still doesn't work out perfectly, but it comes closer. So I suspect that was it.

They left out important details in the paper, plus not being great at long division or rounding. That's our best guess. But I'm still a little suspicious overall of how they dealt with numbers, and that's because of what they reported for that negative control experiment with the male interviewer.

[Kristin]
Oh?

[Regina]
Uh-huh. Okay, so that time they had a sample size of 20 essays for each bridge.

They used the same scale, one to five. They reported average scores of, get ready, Kristin, 0.80 and 0.61 for the two bridges.

[Kristin] (30:28 - 30:39)
That is a really easy math problem, Regina, because if the lowest possible score is a one, how do you get averages that are below one?

[Regina] (30:40 - 30:47)
Some of the stories must have been scored as zero, and I'm thinking maybe someone wrote about celibacy. They subtracted a point.

[Kristin] (30:48 - 31:01)
Could be, but unlike with our smooch scale, Regina, which is just for fun, and we are allowed to make up deviations like martinis in the face, this is a scientific study, and they can't just change their scoring system on a whim.

[Regina] (31:02 - 31:06)
You mean our smooch reading scale is not scientific, Kristin? I'm devastated.

[Kristin] (31:06 - 31:09)
It is trademarked and original, but definitely not scientific.

[Regina] (31:11 - 31:28)
Okay, okay. So, Kristin, by the way, even if they did allow zeros, and even if we allow for each essay to be the average of two whole number scores, even then, 0.61 is still mathematically impossible with a sample size of 20.

[Kristin] (31:28 - 31:34)
Oh, no. So, we have the same GRIM problem again, as well as the averages being out of bounds.

[Regina] (31:34 - 31:40)
Yeah. So, I thought of a name for this. Tell me what you think.

Grim Reaper Math.

[Kristin] (31:41 - 32:16)
I love that, Regina. From now on, I think we should call any spooky arithmetic, you know, where people are doing math in ways that aren't allowed, I think we should call these errors Grim Reaper Math. I love it.

You know, Regina, I see this kind of thing a lot in my statistical sleuthing, and people might think we're being kind of nitpicky. You know, maybe just somebody rounded wrong or made a transcription error or forgot to tell you that they averaged the two raters. But as we've talked about before, when you see one of these Grim Reaper Math errors, it's often just the tip of the iceberg.

[Regina] (32:16 - 32:29)
We had a different word for it rather than iceberg before. In the Vitamin D Episode 1, we called them statistical cockroaches. Where there's one, there's many, and they're gross.

[Kristin] (32:29 - 32:31)
They are super gross.

[Regina] (32:31 - 32:57)
Okay. Let's move on, Kristin, to the statistical test they used because they actually got this test right.

[Kristin]
Yay.

[Regina]
Yay. The test they used was a t-test, and the t-statistic was 3.19, which, as you and I know, is pretty large. And, Kristin, I don't think we've done a full statistical detour on t-tests yet, have we?

[Kristin] (32:57 - 33:26)
No, we haven't, so let's do it.

The t-test is a test for comparing two means, and it helps us distinguish signal from noise. And, Regina, I'm just going to use the situation we have in the paper to help illustrate t-tests. So, for example, if the average sexy score on the scary bridge came out to 5.0 and the average on the safe bridge was 1.0, then the signal would be so big I really wouldn't even need a formal statistical test to help me distinguish signal from background noise.

[Regina] (33:27 - 33:29)
Right. In that case, big signal, very clear.

[Kristin] (33:29 - 33:56)
Exactly. On the other hand, if the results had been mean of 2.5 on the scary bridge, 2.4 on the safe bridge, I can tell the signal is so small that it's not going to exceed the noise unless we have, like, a super gigantic study. But what about the results the authors actually saw in the paper?

2.47 versus 1.41. I'm going to round for simplicity, 2.5 versus 1.4. Here, it's not clear to me whether the signal is bigger than the noise.

[Regina] (33:57 - 34:22)
And that's where a t-test can help. So, here the signal, the average difference, is 2.5 minus 1.4, and that's about 1.1 on this sexy story scale. And that would be, like, going from almost girlfriend to almost kiss.

Yeah. And we can compare this signal to background noise.

[Kristin] (34:22 - 34:27)
And the background noise is measured with a quantity that I love called standard error.

[Regina] (34:27 - 34:38)
You do love it. And by the way, if prospective students want to ace Kristin's Ph.D. program interview, they better bone up on standard errors.

[Kristin]
Don't give that away, Regina.

[Regina]
Oops.

[Kristin] (34:38 - 35:36)
All right. The standard error gets bigger when you sample fewer people or when your outcome, like the sexy scores, is more variable, meaning it bounces around more. So, for example, if the researchers here had rated every essay as a 4 or a 5, that's not a lot of variability.

Whereas if they had a whole bunch of 1s and 2s plus a lot of 4s and 5s, the variability would be quite high. Tiny samples in any of the groups or high variability that wild scatter, all of that inflates the standard error and buries the signal. This is why we like big samples and less variability, because this makes the signal pop more.

The t-statistic that you mentioned earlier, Regina, that was about 3.2, that is a kind of ratio of signal to noise. And usually anything more than a t-statistic of around 2, that's considered statistically significant because that translates to a p-value of less than 0.05. So, definitely a strong signal here. And what was the actual p-value, Regina?

[Regina] (35:36 - 36:15)
The p-value was 0.003, which is highly statistically significant, or, as I'm starting to say now, statistically discernible, because you can discern the signal from the noise. Kristin, a little side trivia here. The t-test was invented by a man named William Gossett who worked at Guinness Brewery in Dublin.

But Guinness didn't let employees publish under their real names because they did not want competitors stealing their beer secrets. So, he published under the pseudonym A. Student. And that's why some books still call it the Student T-Test, which is how I learned it.

[Kristin] (36:15 - 36:18)
That's how I learned it, too, and that's fun trivia, Regina.

[Regina] (36:18 - 36:24)
Right, right. Okay. So, now we've covered results from the erotic fiction out to come.

Right.

[Kristin] (36:24 - 36:35)
So, to sum up, they did find significantly more sexy content in the essays written on the scary bridge compared with the safe bridge. We do have some concerns about the numbers, though, because the means reported are weird.

[Regina] (36:36 - 36:47)
Mm-hmm. Nice summary. Now, let's move on to the next part, the callbacks.

This is where the female interviewer offered the male participants the phone number to talk about the study more.

[Kristin] (36:47 - 36:57)
Right. And then the researchers tracked whether or not the men took the phone number, and of those who took it, whether or not they actually called her back. And we're comparing those numbers on the two bridges.

[Regina] (36:58 - 37:14)
Exactly. Okay, so it turns out men on both bridges took the phone number at roughly equal rates. 78% of men on the scary bridge took her number.

73% on the safe bridge. And that was not statistically significant. The p-value is .76.

[Kristin] (37:14 - 37:23)
So, lots of men took her number, but fear didn't seem to make a difference in this outcome. It doesn't provide evidence of their hypothesis.

[Regina] (37:24 - 37:47)
Right, because I would have thought that if this fear-horniness link is going to show up anywhere, it's going to be in the moment, right? With a guy all excited on the bridge and his heart pounding, and he's like, yes, give me your phone number. But no.

And here's the bad part, Kristin. The investigators just ignored it. No hypothesis test, no p-value.

I had to calculate that p-value myself.

[Kristin] (37:47 - 37:54)
Oh, that .76 p-value, Regina? You just calculated that on your own? It wasn't in the paper?

[Regina]
Yep.

[Kristin]
Interesting that they left it out.

[Regina] (37:56 - 38:01)
Interesting, right? It's one way to handle inconvenient results, right? Just sweep them under the rug.

[Kristin] (38:02 - 38:03)
Yeah.

[Regina] (38:03 - 38:28)
But what about who actually called her back? Right, because this callback rate is really the most important part of the bridge study. And here's what they found.

Nine out of 18 on the scary bridge versus two out of 16 on the safe bridge called back. Oh, that's actually a big difference. Mm-hmm, yep.

They did a chi-square test and got a p-value. Do you want to explain chi-square test really briefly, Kristin?

[Kristin] (38:28 - 38:45)
Sure. The chi-square test compares two categorical variables like did he call or not. So it's also comparing two groups similar to the t-test, but it's about categorical data rather than numerical data. And was the difference statistically significant?

We do have pretty small numbers here.

[Regina] (38:45 - 39:20)
Yeah, but it was. It was statistically significant. They reported p less than .02. Oh. Although, Kristin, fun fact, they listed the wrong chi-square statistic based on the numbers they reported. I don't know. Their arithmetic was wrong again.

They reported 5.7, and really it was a little less. It was 5.4. Oh, so spooky Grim Reaper math again. Maybe they needed a better slide rule.

That's what I'm thinking. Okay, but the correct p-value is .02, which is still significant. So I'm not going to pick on them too much for this one.

[Kristin] (39:20 - 39:34)
A p-value of .02 does sound impressive. But again, this was based on very small numbers. It was a total, what, of 34 men that took the number, and one of the bridges had just two callbacks.

So not very robust.

[Regina] (39:35 - 39:55)
Not robust at all, exactly. Okay, now that is the end of Experiment 1, Kristin. They got significant results, but not on all the outcomes, and the arithmetic is questionable.

Plus, there were those issues with the experimental design, which Experiment 2 tried to address, in part.

[Kristin] (39:55 - 40:02)
Right. Experiment 2, you said, was the one where they only used the scary bridge. They're getting only the same kind of adventurous man, right?

[Regina] (40:03 - 40:11)
Right, right. So this is their attempt to fix that adventure dude confounder, because they didn't randomize men, the bridges, in the first one.

[Kristin] (40:11 - 40:24)
Right, and this is a huge problem, because obviously the guys who chose to get on the swaying rope bridge might have been more thrill-seeking in general, so maybe they would just write sexier stories or call a woman back, no matter what.

[Regina] (40:24 - 41:07)
You know, I can really picture this. I think adventure dude might be a little bit more likely to do that. Okay, so the investigators this time, they kept the setup mostly the same.

Same creative writing task, the same offer of the phone number, but the big difference was when the men got stopped. So some men were approached on the scary bridge, just like before, but others, the control group, were approached afterwards. The investigators said it was at least 10 minutes later, when they were sitting on a bench in the park.

They watched them cross the bridge and then track them down, sitting on a bench in the park, and presumably at that point they were no longer aroused by the bridge.

[Kristin] (41:08 - 41:20)
Okay, so instead of comparing scary bridge guys to safe bridge guys, now they're comparing scary bridge guys to cooled-down scary bridge guys. So same type of adventure-seeking dude, just a different moment.

[Regina] (41:20 - 41:46)
Right, okay, the results. They reported average sexual imagery scores of 2.99 versus 1.92 for groups of size 23 and 22. And I'll just tell you, Kristin, those means are not mathematically perfect, but close enough, so we're just going to move on.

So they did a t-test and they reported a p-value of less than 0.01, highly statistically significant.

[Kristin] (41:46 - 41:55)
So again, guys standing on the scary bridge wrote stories with more sexual imagery than those who were cooled down and sitting on a bench later. Interesting.

[Regina] (41:56 - 42:02)
Better erotic fiction on the scary bridge, yeah. Maybe they ought to have, you know, writing workshops.

[Kristin] (42:03 - 42:08)
I think we should hold a writing workshop, Regina, on the bridge.

[Regina] (42:08 - 42:23)
On the bridge, and then it could be like a little singles mixer, too. Well, I'm talking for scientific manuscripts, though. Okay, we're going to make this work, Kristin.

We're going to trademark this again. Oh, wait, reality show. It's just going to be a reality show.

[Kristin] (42:24 - 42:34)
I'm thinking scientific papers are the most boring things ever to read. So if we got people out on the sexy bridge to write these papers, then maybe they would be more exciting and more people would read about science.

[Regina] (42:34 - 42:46)
I like it. Let's go for it. Okay, so back to what they did, better erotic fiction on the bridge, but their reported results for this section used the wrong sample sizes.

[Kristin] (42:46 - 42:46)
Oh, no.

[Regina] (42:47 - 42:56)
So I had to fix that. Now, their error, again, did not change the conclusion, but Kristin, can I just say they're not exactly going to win any awards for best arithmetic?

[Kristin] (42:56 - 43:05)
Yeah, I'm really wondering, did they just, you know, do this in a notebook and the researchers couldn't add? All right, but Regina, what about the phone number stuff?

[Regina] (43:05 - 43:46)
Phone number stuff. Okay, first of all, as before, no significant difference in how many men took the phone number from her. And again, the investigators conveniently ignored that.

Now, as far as who called her back, the authors messed up the numbers, but this time, Kristin, it was a more serious mistake.

[Kristin]
Oh, what do you mean?

[Regina]
So I'm going to frame it as another quiz for you, Kristin, see if you can spot the issue.

The authors report that in the experimental group, 20 out of 25 accepted the phone number and 13 out of 20 called her. And in the control group, they said 19 out of 23 took her phone number and 7 out of 23 called.

[Kristin] (43:47 - 43:59)
Okay, wait a minute. You said 19 took her phone number, but 7 out of 23 called. It should be, though, 7 out of 19 called because only 19 took her phone number.

So they used the wrong denominator.

[Regina] (43:59 - 44:34)
I have kind of a basic math problem here, and this time it turns out to be important. So they reported a statistically significant p-value of less than .02 for their chi-square test. But when you fix the numbers with the correct denominator and rerun the analysis, the p-value goes up to .08. Which is not statistically significant. No, but it means their big result disappeared. So they cannot actually say that more men on the scary bridge were more likely to call her back in experiment two.

[Kristin] (44:35 - 44:40)
So in this experiment, the only positive results were from the sexy essay then.

[Regina] (44:40 - 44:59)
Mm-hmm. It's kind of had the weird numbers and the weird averages. So I'm feeling like shaky results.

Shaky bridge, shaky results. Yes. That's what it is.

Okay, now the third experiment, no bridges this time. This one's in the lab, and there's electric shocks.

[Kristin] (44:59 - 45:41)
Oh, sounds fascinating, Regina. I can't wait to hear, but let's take a short break first.

Welcome back to Normal Curves.

Today we're asking the important question, does fear make you horny? We've already walked through the famous scary bridge study and all its problems. But, Regina, you promised us another experiment in a new setting and some possible replications.

[Regina] (45:42 - 46:02)
Yep. Now, Kristin, this last experiment was a little confusing, so I'm just going to hit the highlights, okay? So there were 80 male university students in the lab, and the basic idea was this.

Let's make half of them believe they're about to be painfully shocked with electricity.

[Kristin] (46:02 - 46:02)
Oh.

[Regina] (46:03 - 46:40)
Yep. The setup was to have one man in the lab at a time with a pretty woman, of course. And the pretty woman was supposedly another participant, but she was actually a confederate in the experiment.

Ah. And the researcher told the pair that a coin flip would decide whether they'd get a painful shock or just what they called, quote, a mere tingle. In fact, some subjects describe it as enjoyable, which, Kristin, actually makes it sound kind of sexy, right?

A mere tingle, enjoyable, and it's making me think of vibrators, honestly.

[Kristin] (46:41 - 46:44)
I'm not going to talk about vibrators on this podcast, Regina.

[Regina] (46:45 - 46:49)
Not yet, you're not. Never say never.

[Kristin] (46:49 - 46:58)
Okay, so now the fear comes from anticipating a painful electric shock. I'm assuming they didn't actually get a shock in the end.

[Regina] (46:58 - 47:04)
They did not. It was just the threat of one that was supposed to make them aroused enough.

[Kristin] (47:04 - 47:09)
Ah. And the research question is, does the fear make the woman nearby seem sexier?

[Regina] (47:10 - 47:56)
Exactly. So they had a questionnaire that asked the man about this directly. How much would you like to ask that woman out on a date, and how much would you like to kiss her?

On a one-to-five scale. Yep.

[Kristin]
Wow, straight to the point.

[Regina]
Straight to the point. There was no beating around the bush like they did with the bridges. Now, they had a cover story for why they were asking the guys if they were hot for their co-participants, but it still seems like kind of a loaded question.

Oh, by the way, how hot are you for this random woman sitting next to you? We swear, it's not the point of the experiment itself. No, no.

Right, right. Nothing to see here. Okay.

And what they found was that guys who thought they were about to be painfully shocked were much more attracted to the woman next to them. Highly statistically significant.

[Kristin] (47:56 - 47:58)
Oh, interesting. And this one's randomized, actually.

[Regina] (47:59 - 48:09)
I know, right? They actually did it in the lab. This was controlled.

And just like with the bridge study, though, they had them do that same creative writing assignment again.

[Kristin] (48:10 - 48:11)
They really loved that test.

[Regina] (48:11 - 48:29)
They did the erotic fiction. Okay, and the results for that one were mixed and kind of strange, actually. They only reported subgroup results, not the full group results, and they said that only one subgroup was significant.

[Kristin] (48:30 - 48:43)
Oh, no. This is kind of reminding me, Regina, of the pheromone study where they reported only some subgroups with weird analyses, and when you need to go hunting in subgroups like this, it's a bit of a red flag.

[Regina] (48:44 - 49:02)
Yeah, and Kristin, again, the numbers were strange. The average scores didn't make complete sense. And what's even worse, their reported statistics did not line up with their reported p-value, and so I can't tell.

Was it even statistically significant or not? I have no idea.

[Kristin] (49:02 - 49:06)
Oh, weird. So we're back in the statistical twilight zone.

[Regina] (49:07 - 49:08)
Mm-hmm, exactly.

[Kristin] (49:08 - 49:13)
So a lot of mixed results here, but I want to hear about the replications that you mentioned earlier.

[Regina] (49:14 - 49:34)
Yeah, so just a few years after the scary bridge study was published, Kristin, it was already common in psychology textbooks and in the popular imagination, really. But as far as I can tell, no one has ever tried to recreate the bridge setup or even anything remotely like it.

[Kristin] (49:34 - 49:41)
Well, that makes sense. People want to do replications, but you really can't replicate this bridge study. It wouldn't be ethical.

[Regina] (49:42 - 50:12)
Kind of lived on in infamy and will forevermore. But other researchers did try to replicate the ideas behind it, including that electric shock experiment that we just talked about, and that was actually the subject of one of my favorite replication attempts. It was from 1979, another team, and their paper was titled Misattribution Under Fear-Producing Circumstances, Colon, Four Failures to Replicate.

[Kristin] (50:13 - 50:20)
Wow, that's a very direct title. Four failures? So they tried it four times, and all the replications failed.

[Regina] (50:20 - 50:55)
Yep, and they got kind of salty about it. You can read it in their paper. They said they had set out on kind of a routine replicate and extend.

So they didn't really question those original shock experiment results. They thought that they were solid. So first they did a conceptual replication, expecting it to come out, and then they were going to go further, but that conceptual replication did not replicate.

So they made their design even more like the original study's design, you know, thinking that was it, nothing, and then again and again, no effect.

[Kristin] (50:56 - 51:02)
Wow. So they even kept trying to make it more and more similar just to see if they could eke out that result, and nothing.

[Regina] (51:03 - 51:32)
Nothing. And by the end, they were really disillusioned, and they were so disillusioned, they weren't really even buying into this whole theory anymore. And, Kristin, you'll like this, they tried to warn the field.

They wrote, given the failures to replicate and the theoretical problems with this whole idea, social psychologists might well be cautioned to reevaluate their current enthusiasm for publicizing this theory to the lay community.

[Kristin] (51:32 - 51:33)
Wow. That's kind of an academic burn, and good for them.

[Regina] (51:33 - 52:03)
Right? It's like, guys, knock it off.

Quit hyping this. Yeah. Yeah, but of course the warning was ignored.

[Kristin]
Of course.

[Regina]
Yeah, of course. So other investigators then went on to try to replicate the general idea with setups that were a bit safer, including the electric shocks, but other things like scary movies and roller coasters, even listening to descriptions of grisly murders.

[Kristin] (52:04 - 52:05)
Oh, eww. That was one of them.

[Regina] (52:05 - 52:06)
Eww, I know.

[Kristin] (52:07 - 52:09)
I don't know if that's scary so much as disgusting.

[Regina] (52:10 - 52:23)
Disgusting, I guess it does that whole fear thing. I don't know. Anyway, a team of investigators then put together a meta-analysis on this.

They pulled together 33 studies with over 1,200 people.

[Kristin] (52:24 - 52:31)
A meta-analysis. That's fantastic because now we have a large sample, and maybe we can start to get some answers. And what did they find, Regina?

[Regina] (52:31 - 52:38)
Well, first of all, I should say, Kristin, here they pulled together men looking at women and women looking at men.

[Kristin] (52:38 - 52:50)
Oh, this is great. So some of those subsequent studies actually tested whether this worked for women. So women on the roller coaster seeing if the man next to them seemed sexier.

Yes?

[Regina] (52:50 - 52:54)
Uh-huh. Yep. Women's lib finally got caught up.

[Kristin] (52:54 - 52:54)
Finally.

[Regina] (52:55 - 53:18)
Overall, they found that being scared, yes, did increase horniness, but just a tiny bit. A tiny bit, but it was still statistically discernible.

[Kristin]
Oh.

[Regina]
So when people were aroused by fear, they found the other person more romantically attractive, more physically attractive, and also more likable. Again, slightly, but it was there.

[Kristin] (53:18 - 53:23)
So even though the effects are small, there still might be something there. Wow.

[Regina] (53:23 - 53:58)
Uh-huh. But wait, Kristin, it's not the full story. There's a twist.

Okay. And I found this next part really fascinating. Now, full disclosure, they also included studies that used arousal from things other than fear.

But it's plausible that it applied to fear alone, so I'm just going to report it like that. Okay. Okay.

So here is the twist, though. Fear only increased horniness and attractiveness and likability if the other person was already objectively very good-looking. And even then, it was still only a small boost.

[Kristin] (53:58 - 54:10)
Regina, it sounds an awful lot like the red dress effect, back in the red dress episode. It only works if you're already attractive, and even then the effects are very small.

[Regina] (54:11 - 54:58)
Yeah. But, okay, here's another twist that we did not find in the red dress effect that I found a little depressing, actually. What if the person was the opposite of a hottie?

What if they were, oh, let's just say ugly?

[Kristin]
Oh.

[Regina]
That's what they called them in the meta-analysis.

It ranged from ugly to beautiful.

[Kristin]
Wow.

[Regina]
I know.

They were right out there. Okay, if the person was a non-hottie, an ugly person, then fear could actually make things much worse for them. So if you're around an unattractive person and you're scared, then it becomes a negative effect.

Now not only are they not more attractive and more likable, they're actually way less attractive.

[Kristin] (54:59 - 54:59)
Oh, wow.

[Regina] (54:59 - 55:46)
Way less likable, yes. And these negative effects, by the way, disappointingly enough, were even stronger than the positive effects for the beautiful people.

[Kristin]
Wow.

So this trick could backfire is what you're saying, Regina.

[Regina]
I know, right? Like you got to be worried about these little psychological tricks.

They don't apply to everyone in every situation. But there is a silver lining, Kristin. What about, for the rest of us, normal, average-looking people?

Okay, fear didn't do much, but at least it didn't make things worse. Nice. It might give you a little bit of a boost on physical attractiveness.

Again, not huge. But at least it's not going to turn you into a goblin, you know?

[Kristin] (55:46 - 55:56)
Well, that is good news for the rest of us, Regina. But you're saying if they're hot, then they should put on a red dress and go to a haunted house.

[Regina] (55:58 - 56:09)
What could they dress up as, someone in a red dress, for Halloween?

[Kristin]
Devil.

[Regina]
Oh, devil.

Oh, I wonder if that's why everyone loves this sexy devil costume.

[Kristin] (56:09 - 56:10)
Maybe, yeah.

[Regina] (56:11 - 56:36)
And it might work for both men and women. I was thinking the red dress you could go as that emoji. You know, the woman in the red dress dancing, that emoji?

[Kristin]
I don't know that emoji.

[Regina]
There's an emoji with a woman dancing in a red dress.

[Kristin]
Okay.

[Regina]
You'll go as the devil, though. I think you would make a very nice devil. Okay, let's recap, then.

[Kristin] (56:37 - 57:13)
Right. So the original Scary Bridge study found some evidence of the claim we're looking at today, but it wasn't on all the outcomes, and the math was questionable, and some of the numbers were really small, so it didn't seem totally robust. And then some replication attempts failed to find anything at all.

Finally, a meta-analysis of a bunch of subsequent replication attempts did find that fear, indeed, might make you horny, but the effect was small, and it was only in a subject that was already attractive, and it might even backfire for someone who was, as you called it, a non-hottie, Regina. So, mixed bag.

[Regina] (57:14 - 57:23)
Mixed bag. I love how it's not quite the simple story that everyone makes it out to be when they talk about the study. Yeah.

Not quite so simple.

[Kristin] (57:24 - 57:30)
All right, Regina, I think it's time to wrap things up and evaluate our claim for today. So, Regina, repeat the claim.

[Regina] (57:30 - 57:37)
Right. It's that heart-pounding fear makes men see women as more attractive than they would be otherwise.

[Kristin] (57:38 - 58:01)
And how do we rate the strength of the evidence for this claim? It's with our smooch rating scale, where one smooch means little to no evidence, and five means very strong evidence. But, Regina, I promised earlier in the episode that just for today, we are going to change our rating scale to the sexy erotic fiction scale.

And remind us of how that scale worked, Regina.

[Regina] (58:01 - 58:27)
So, one meant no sexual content, so let's say chastity. Two meant they used the word girlfriend or the equivalent. Three meant they used the word kiss.

Four meant they used the word lover. And five meant they talked directly about the deed about sexual intercourse. So, we've got chastity, girlfriend, kiss, lover, or sexual intercourse.

[Kristin] (58:27 - 58:32)
So, Regina, how are you rating the evidence today on this temporary scale of ours?

[Regina] (58:33 - 59:45)
Yeah, so I'm going to give it two, a girlfriend. Or maybe a boyfriend, a significant other. I think the sexy bridge study itself was really flawed, but also it was charming.

Those were different days, and there were no ethics or safety review boards, and you could just assume there was sexual undercurrent to everything. And numbers don't add up, and no one notices or cares. I think it was not a strong study in its own charming, wonderful way.

And I think that any effect from fear is going to be small. And I can believe that it works only for already great-looking people. So, I thought about bumping this up to a three because it was the 1970s.

So, I thought about bumping it up to a kiss. I couldn't find any preregistered studies on this, though. So, I'm still waiting for more firm evidence, and I don't think I'm ready to kiss the scary bridge or take it on as a lover or have sexual intercourse with it.

What about you? What are you willing to do with the scary bridge?

[Kristin] (59:47 - 1:01:20)
Regina, so you're saying none of the studies in the meta-analysis were preregistered?

[Regina]
No, they were not as far as I could tell.

[Kristin]
Yeah, good to know.

So, Regina, I'm actually going to go 2.5. Somewhere between a boyfriend and a kiss. And, you know, again, it's the charming, classic study. I do want to give the authors some credit because I do think that they thought through with some common sense.

They had that negative control of the male interviewer. They also did experiment 2 where they had only scary bridge men. And I think that took a lot of thought, actually, even if their study was imperfect.

They tried, at least. And I'm going to give them, you know, points for effort here to address design flaws. The Grim Reaper spooky math, that really worries me when I see that kind of problem with the math.

I'm not going to take off as many points here, though, because I'm assuming that this math was done by hand. And so I don't think they did something underhanded like you might do when you have a computer, which is run everything a million different ways, throw people out, right, see how you can get the right result. That would be hard to do by hand.

So I'm wondering if maybe they just weren't the best with arithmetic or there's some silly explanation like that. So I'm not going to hold it against them quite as much as I would with a modern study. And there are some replications supporting that maybe there is some effect here.

So, yeah, I'm going to go with 2.5.

[Regina] (1:01:20 - 1:01:24)
Hmm, I don't know. Maybe we should round you up to a 3 using some spooky Grim Reaper math.

[Kristin] (1:01:24 - 1:01:27)
I can't kiss it because of the math, Regina.

[Regina] (1:01:30 - 1:01:37)
Maybe just like an air kiss. What about your methodological moral? Do you have anything good there?

[Kristin] (1:01:37 - 1:01:49)
Yeah, I've got a good one for today. I am focusing on the Grim Reaper math idea. So here's mine.

Those who don't verify their numbers dig their own statistical graves.

[Regina] (1:01:51 - 1:02:33)
Oh, that's nice and perfect for Halloween too. I went with the Halloween theme, yes. Spooky.

Oh, I like it.

[Kristin]
How about you, Regina?

[Regina]
So I want to do something about the old study but a famous study because sometimes I think people give these classic famous studies like a free pass.

They don't actually pull the study and read the study. So here's mine. Famous doesn't mean flawless.

[Kristin]
And, Regina, that actually has applications outside of studies. You and I know some famous people. Who are not flawless.

[Regina]
Who are not flawless, let's just say that. Kristin, this has been a lot of fun, and I just want to say Halloween's coming up. I definitely want to go for a haunted house.

[Kristin] (1:02:34 - 1:02:34)
You should do it, Regina.

[Regina] (1:02:35 - 1:02:55)
I should. I don't know about the, you know, sexy devil. Maybe I'll just dress up as a bridge.

Can I do that?

[Kristin]
Oh, Golden Gate Bridge, it's red.

[Regina]
Oh, that's perfect.

Okay, we might have to work on my costume, but I'm going to go with the Golden Gate Bridge. Okay.

[Kristin] (1:02:56 - 1:03:01)
All right, Regina, this has been a lot of fun. This has been super interesting. Thanks so much.

Thanks, Kristin.

[Regina] (1:03:02 - 1:03:03)
And thanks, everyone, for listening.