Results of the survey on Dark Secrets, part one

A little while ago via this blog, I released a survey in which I asked people to answer some questions relating to dark secrets and their experience of dark secrets. You can still take the survey here if you like, though I’ll be analyzing in this the first 150 responses, future analysis might include any new responses.

I’m going to be releasing the results of this survey in two tranches. This is the first tranche and will consist in results and analysis of numbers and proportions of answers to questions considered on their own. The next tranche will consist of the analysis of correlations, odds ratios, and other relationships between answers to different questions. This is because I’m going to have to do a lot of coding to get the data set up properly for the latter kind of analysis.

1. Number of responses

The first interesting result to report is that it appears a large number of people, proportional to the number of people who saw the survey shared somewhere, were willing to fill it out. I have experience collecting responses on surveys, and it wasn’t notably difficult to get people to respond to this one- the ratio of shares to responses was about the same, if anything it might have been a little better. Given how unpleasant a topic this is to think about, and given understandable concerns about results leaking, this might be a surprise to some. I suspect that the inherent interest in the topic counterbalanced these barriers. Speculating a little, I think that keeping secrets is a lonely and tiresome business, many desire to let their guard down- if only a little- and these sorts of people were likely attracted to this survey.

2. How common are dark secrets?

I defined a dark secret as follows:

“By "dark secret" in the context of this survey I mean something special. I mean a secret about you, something you did or failed to do, which, if revealed, could severely harm your reputation with your peers. A dark secret doesn't have to be objectively bad, it just has to be something which, in your view, would harm your reputation given the context of your life. Homosexuality and gender non-conformity aren't wrong (at least in the survey author's view) but for some people would count as a dark secret. Do you understand?”

Only 7% of respondents said that they didn’t have a dark secret. An additional 14% said they probably didn’t have one. 24% said that they probably did have a dark secret but weren’t sure while 58% unequivocally said they did. Granted, this has got to be due, at least partially, to selection effects on who chose to fill out the survey. Still, I suspect these results probably are at least very broadly reflective of numbers from the general community of the educated, online set who congregate at places where I shared this survey, like the Slate Star Codex Subreddit, this blog, and the sample size Subreddit. The extent to which they are reflective of the general population is something I won’t hazard a guess on. I suspect the substantial majority of adults do have dark secrets, though that’s based more on a general impression than the results of this survey.

3.1. Severity of secrets- How serious do people think their own dark secrets are in comparison to everyone else’s

A result that grabbed my eye was that only about a quarter of people thought that their dark secrets were more serious than other people’s dark secrets. By comparison, about half of respondents thought that their secrets were less serious than those of other people. About a quarter thought their secrets were roughly as grievous as average.

The most obvious interpretation of these results is that people underestimate how bad their own dark secrets are or overestimate how grievous the dark secrets of other people are, or both. Another interpretation is that our respondents are correct because the dark secrets of respondents to this survey were on average less grievous than the dark secrets of the whole population.

Another possible interpretation. Assume that the severity of dark secrets has a long tail and is right-skewed. If people interpreted this question as being about their dark secret’s severity versus the mean average severity of dark secrets (though maybe not in such explicit mathematical terms), then this result could make sense. A substantial majority could be below average (that is below the mean, not below the median). This is true even if we assume that the sample is representative of the population as a whole.

A promising avenue for future research is investigating whether people underestimate the relative gravity of their dark secrets. For example, we could investigate this by having people share their secrets with the researchers, getting participants to rate the severity of their own secrets, and then having other participants rate each other’s secrets (names withheld, of course).

3.2. Severity of secrets- How serious are these dark secrets in an absolute sense?

How serious are the secrets we’re talking about here? I think in this regard the most revealing figure is that 61% of respondents thought they had a secret that, if revealed, would cause at least some of their friends to reduce contact with them. 34% of people thought that they had a secret bad enough that it would cost them at least half of their friendships. 12%, nearly one-eighth of respondents, felt that they had secrets so bad that if they were revealed they would have few, if any, friends left.

Now it’s very possible that respondents are overestimating the gravity of their dark secrets, or how quick their friends would be to abandon them. It’s also very possible that some respondents are underestimating how seriously their dark secrets would be received. We also don’t know how representative our sample is. If, however, one-third of the population - or at least of the population you are likely to meet online- is carrying around a secret that would indeed cause half their friends to abandon them, that is an extraordinary and sad thing.

3.3. Severity of secrets- Distress

I was heartened to see that only a little over 10% of respondents reported that their dark secrets caused them severe or very severe distress. Another 16% reported moderate distress while 49% reported slight distress. In a survey filled with troubling results, this was not as bad as I thought it would be.

Still, in the scheme of things, 10% of people experiencing severe or very severe distress is actually a lot of distress.

Am I wrong to feel sorry for these guilt sufferers? Perhaps I should be focusing my sympathy on their victims. Perhaps I should actually want people to be experiencing more guilt. But there is something peculiarly useless about guilt as an emotion. I’m always reminded of something I saw on a Magic the Gathering Card once:

Those who most feel guilt don't need to, while those who most need to feel guilt never do.

Hence guilt is a peculiarly useless emotion in practice. A short sharp stab of it can have its uses, but the long guilt of years seems to lead to nowhere good.

4. The assessed probability of secrets coming out is mostly lowish, but not that low

We asked participants the following question:

“Think of the dark secret you are most afraid of coming out. What do you think the chances are that one day it will come out? Answer as a probability between zero and one, like 0.7, 0.35, etc. If you don't have a dark secret, don't answer this question.”

The median answer was 10%. The 75th percentile thought there was a one-quarter chance of their secret coming out. About 80% of people were in the interval between a 1% chance of their secret coming out and a 30% chance. On the whole, then, people thought it was a real, but fairly unlikely, possibility.

I discarded text answers to this question because they were difficult to interpret. However we had several “it already did” and “in a sense, it kinda already did” type responses.

5. Willingness to pay to protect one’s reputation

Participants were asked

“Suppose you were offered a deal by a genie. You could give a certain number of years off your life, and in exchange, your reputation would be guaranteed forever- against damage from dark secrets and everything else from your past. Transgressions in the future could still harm your reputation. What is the maximum number of years you'd be willing to give? Fractions, e.g. a quarter of a year, are acceptable.”

Years were chosen as the currency because A) It is constant from country to country B) The number of years people have to spare- especially among people who filled out our survey who I assume tend to be youngish- varies less than income does. One survey taker’s year is much closer to being interchangeable with another survey taker’s year, on average, than a survey taker’s dollar is interchangeable with another survey taker’s dollar. In hindsight though, years have another problem- it is taboo in our society to trade life for something else.

The median amount of time people was willing to give was nothing. About the 75th percentile was the borderline between 1 and 2 years. The mean, not including the two people who they said they would give all their years, was about 3 years.

20 out of 141 who answered the question said they would give at least five years. Two people said however many years they had, they would give all of them. The overall picture here is huge variance with a big rightward skew. It will be interesting to analyze this in relation to other variables.

6. Horizontal sympathy

Forty percent of people stated that their own dark secrets made them more sympathetic to others who are outed as having dark secrets. This question was of great interest to me. I have a form of OCD that often makes me think, wrongly, that I have severe dark secrets, and makes me worry disproportionately about those dark secrets I do have. I have observed that this form of OCD has made me more sympathetic to people who actually do have dark secrets and get found out. I was curious to see whether this was other people’s experience, and in a way, cheered that it was.

7. The content of dark secrets

What about the content of people’s dark secrets? Here’s some information on the responses we got, keep in mind that response categories were non-exclusive- you could check as many as apply. There was a short custom answer option, but I will not be reporting on any of those results due to the risk of identification.

A lot of them were sexual. The most popular category was “sexual misbehavior not elsewhere classified” with 44 responses. I was disappointed at how popular this category was because I thought that I had set up enough sexual subcategories that we would be able to classify the majority of sexual secrets. Categories in the survey related to sexuality included consensual homosexual activity (5 respondents), cheating on your partner (14 respondents), sexual fetishes between consenting adults (22 responses), incest, whether consensual or not (7 responses) and “Breaches of sexual consent broadly construed including but not limited to harassment, assault, stalking, statutory rape, rape etc.” (19 responses). If you have any theories on what kind of acts might be going into sexual misbehavior not elsewhere classified I’d like to hear them. I can think of a few things that would fit, but nothing that would explain why it was so overwhelmingly popular. I suspect that many responses perhaps did belong to another category, but people didn’t want to admit it to themselves.

Secrets about gender non-conformity (5) were a fair bit rarer than sexual secrets.

Prejudice of various forms was relatively popular- eight responses for racism, ten responses for sexism, and seven responses for other forms of prejudice.

Miscellaneous “destructive behavior” including physical violence (10), taking something you shouldn’t have (13), and destroying an item you shouldn’t have (9) were common.

One of the most popular categories was “Betrayal of trust” with 28 responses.

Family mistreatment other than your own children or partner (3), mistreatment of a partner (4), and mistreatment of your own children (3) were rare. I have a sense that a lot of people may be blind to their own abusive behavior, so I suspect the real figures are higher.

Incompetence (13) and negligence (14) were also quite common.

8. Gender demographics

One thing that will limit the generalisability of our findings- at least beyond these kinds of internet spaces- is the small number of women who responded- only about 17%. Prefer not to say and other made up 8% of respondents. The rest were men.

9. A bimodal distribution of happiness

Something that was very interesting to me, although it is not directly related to the subject at hand, is that we found a bimodal distribution of happiness, with one centre at 3 and another at 7 (happily the centre at 7 was somewhat larger than the one at 3. What gives? Well, it could be a feature of the population we sampled, but I do not think so. I think it was a feature of the unusual question I was trialing:

“On a scale from 0 to 10, where 0 is the saddest person you know, and 10 is the happiest person you know, how happy are you?”

My theory is that people looked at this, and tended to immediately have one of two reactions “I’m not the saddest person I know, but I’m sadder than most” or “I’m not the happiest person I know, but I’m happier than most”. These respectively corresponded to a centre of gravity of 3 or 7. Why there weren’t more people who immediately thought “I’m about typical for the people I know” is a mystery to me.

10.1. Concluding thoughts- My suspicions about dark secrets in the population

I first became interested in dark secrets, beyond the context of my own strange OCD obsessions, when about two years after metoo I started counting and I realized that I personally knew well over thirty people who had been accused of intimate partner abuse, sexual harassment, or sexual assault, with varying degrees of severity. Now it must be said that I know a lot of people, and the people I know are of a set who are unusually likely to use (sometimes dueling) public accusations to try and punish perceived offenses. Nevertheless, that’s a lot of dark secrets.

My strong suspicion is that a majority of people have, at some point in their life, done something really bad. We’ll operationalize “really bad” vaguely as “at least as bad, in the public view, as having had an affair while married”. Doubtless, many people don’t recognize their own misdeeds. Their brain won’t let them clearly perceive what they’ve done. Others know it, but wouldn’t admit it, even in an anonymous survey. But I would confidently guess that at least fifty percent of adults have, at some point, done something “really bad” on this definition (again- greater than or equal to having an affair while married).

But having an affair is only the minimum, many people have done much worse. Going back to my initial point in this section, consider how common perpetrating sexual and domestic violence must be. 20% of women report having been raped, or subject to attempted rape. One in four women in the US report being the victim of severe physical violence by an intimate partner. Many, many more, probably solid a majority, have been subject to some form of sexual violence or domestic abuse, including groping, stalking, emotional abuse, and harassment. Indeed if we use a broad but not exceedingly broad definition of sexual and domestic violence, we’ll probably find that most men have been subject to sexual or domestic violence too. I know I have. Now even granting that most perpetrators are repeat offenders, that means there’s got to be a fair chunk of offenders out there. As I mentioned earlier, 19 respondents to our survey admitted to “breaches of sexual consent broadly construed including sexual harassment, sexual assault, stalking, rape or statutory rape”. How many more have done it, but didn’t want to admit to it in a survey, or haven’t acknowledged the nature of their own actions to themselves? I’m sure there are graduations of severity here, and probably most of those 19 were towards the low end, but we’re still talking some pretty dark secrets here.

What else is there that’s seen as at least as severe as having an affair that can be kept as a secret? Lots of things I suppose. Extreme cruelty to animals, gross betrayal of friendship, destruction of a person’s reputation in secret, embezzlement, and racist abuse. Even keeping a dark secret can be a dark secret- how many people out there know of a terrible crime yet do not report it and are eaten up inside by that?

10.2-Concluding thoughts- how moral differences can conceal dark secrets even from their bearers

Sometimes people don’t even know that they have dark secrets due to contextual moral differences. There are many examples of this but my favorite is Grindr dick pics. Recently the United Kingdom has decided to outlaw the practice of sending dick pics. A (male) journalist in response to this tweeted that he’d been sent several hundred naked photos that he hadn’t asked for, and that while it could be irksome, he really didn’t think all those people should go to prison. A bunch of people wrote back to him calling an egotistical liar, saying no man would get that many, and he was clearly lying probably to defend his own practice of sending dick pics. The dullards abusing him either didn’t realize he was gay or didn’t know that this is completely standard for gay men. I can confirm this, I’ve received hundreds of unsolicited dick pics myself.

Now in the minds of the furious people, what these guys have done in sending me and this journo dick pics is unconscionable. A half step below flashing someone on the street, if that. But the interesting thing about the feelings of these “cyber-flashers” is not that they disagree about this it’s that they’re not even aware anything they’re doing would be controversial. To them, it’s just a commonplace, not worth remarking on. Indeed to remark on it is a bit weird. I once had a guy ask me for pics and I wrote back “does that include dick pics?” he wrote back with something to the effect of an eye-rolling emoji and “duh, of course”. He found it strange that I would need to clarify that he wanted dick pics- why not just send them straight up when asked for pics? All these Grindr denizens holding a secret that could get them in trouble if it were publicly revealed, and not even known. For them, it’s just natural, a courtesy like showing a prospective buyer a sample of wares.

But the difference between moral contexts cuts deep in a really interesting way. I bet a lot of these very same men if they read an article in the paper about some social media sex pest sending women dick pics would tut-tut and wouldn’t even have the thought that there was an inconsistency there. And in a way, they’d be right, because there is a huge difference in the meaning of that act between those contexts, but it’s not necessarily a distinction the larger public would acknowledge or accept if these Grindr users’ past behavior became public. In the UK for example, it's not as if this cyber-flashers bill has a Grindr exception.

10.3.- Concluding thoughts- Weird darkness

I bet a lot of dark secrets are just weird - random acts of evil are more common than we sometimes realize. I remember I was walking down the street once when I saw a woman on a hen’s night start kicking a homeless person and laughing maniacally. I chased after her but I wasn’t fast enough to catch her. The homeless guy said he had no idea why she’d done it.

10.4.- Concluding thoughts- why I couldn’t ask the mos interesting question f them all

I’m sad that I couldn’t bring myself to include a question in which I just outright asked people to describe what they’d done. My reason for not doing so? It’s possible that, in the future, AI-based stylistic analysis of writing, combined perhaps with a few other clues, will be able to identify individuals based on passages they’ve written. Moreover, people tend to give away clues about their identity in their writing, even when they don’t mean to. Thus I could not justify it to myself from an ethical perspective.

10.5.- Concluding thoughts- Luck in who faces consequences, and moral luck in general

We do and see these terrible things, and yet only a minority of us will face genuine consequences, and the distribution of consequences seems random- at least from the point of view of justice. It sometimes seems to me that we think- at least unconsciously- that so long as you don’t get caught for something it’s not so bad. Sometimes we even codify it. I once saw a woman express the view that if you married a “real alpha” he was bound to cheat, but if he truly loved you he’d make sure you never found out. Granted, sometimes getting caught can be a signal of special recklessness, or that you’re more likely to have caught because you did it many times, but our habit of harshly socially punishing people for things that are relatively common, but are just rarely found, doesn’t seem right to me. Still, what’s the alternative? There’s some really awful shit going on out there, and if we can’t consistently catch the people who do it, arguably that’s a reason to amp up the punishments so the limited number we do catch can be used as strong deterrents.

Obiter dicta, all of this puts me in mind of the conceit of moral luck. You run over and kill a pedestrian while speeding at 25%+ plus the speed limit? That’s a huge deal and is thought to reflect very badly upon you- a time in jail is probably expected, even for a first offense, at least in the USA. Even if you don’t go to jail, you’re expected to feel bad about it every day for the rest of your life. You go 25%+ the speed limit and don’t run over anyone? Ahh man, we all make mistakes, don’t worry about it bud, the speed limit should be higher anyway. It’s certainly not a dark secret.

10.6- Concluding thoughts- Is doing the wrong thing something you can just fall into?

There’s a character in a video game I’ve played -Wrath of the Righteous- called Regill. In one of the most quoted lines of the game, he says that he’s sick of hearing excuses:

Every betrayer has their own sob story to excuse their actions. And each one thinks they are different from the rest, that they alone should be understood and shown mercy. I'm sick of hearing it.

At the end of the day though, Regill would maintain it’s pretty easy not to throw your lot in with cosmic evil. When I was talking about this survey with a work colleague, she expressed pretty much the same point of view about people who cheat on spouses they claim to love: it’s easy not to cheat. In the process of cheating, there are dozens of points at which you could stop yourself. Forget this forgiveness crap, this isn’t an accident.

We could say the same thing about most awful, reputation-destroying acts I guess. There’s the odd exception- the Grindr case we discussed, or Aziz Ansari types who (without any comment by me on Ansari’s particular case which I have not thoroughly investigated) are thought to have made bona fide mistakes and paid for them. Doubtless, there are a few other cases at least bordering the thicket of moral uncertainty. For the most part though, to do something seen as really bad, you’ve got to make a deliberate choice.

Yet though this seems to be true a priori, in practice the species keeps making mistakes and I do not think it’s because we are mustache-twirling villains. I think bad people are rare, but people who have done really bad things are not. How are these reconciled? I don’t know.

As you look around you and quantify the probable sins of your fellows and yourself how should you feel? When I first started thinking about this years ago, I framed it as a kind of free existential choice. There’s no level of sin at which, logically, you must stop loving someone, so that leads you to a kind of unbounded decision, a kind of ethical aesthetic. Will you be on the side of humanity, warts and all, or not? Is there a way to keep loving a species riddled with dark secrets without excusing them?