The Stepford Wivesbands Prison Experiment
October 3, 2010 5:01 AM   Subscribe

What would be the best empirical and scientifically grounded experiments one could perform on a potential marriage partner?

Marriage is one of the riskiest ventures one can make in life; however, it seems that it is also one of the least objectively researched endeavors. Marriage is time-consuming and has major financial and emotional risks that can be absolutely devastating if it comes to divorce--or worse, yet we seem to put more objective research in other areas such as rigorous personality tests for new employees, reviewing data on automobiles, price checking houses and cars followed by third-party inspections, etc.

We know that we can be blinded by love and sex; however, we push that aside and rely on our flawed intuition.

I know that psychological research is new and that there are historical and cultural reasons as to why we do not place more emphasis on objective data on potential marriages; however, I am wondering which would be the best types of testing that one could do on a potential mate.

I am looking for solid scientific research such as personality tests, behavioural experiments, and anything else that could be objectively quantified. Perhaps tests used by marriage and relationship therapists. Another example could be something like increasing room temperature while the future hubby performs a mental task to see how they perform under pressure.

I am looking for both innocent ideas and more devious ideas that could be shocking or even questionably unethical.
posted by Knigel to Science & Nature (38 answers total) 14 users marked this as a favorite
 
Show them this question and ask them to rate their reaction on a scale of 1 to 10?

(Serious answer)
posted by vincele at 5:15 AM on October 3, 2010 [10 favorites]


There's no need to perform any form of devious or unethical experiments. John Gottman's work has looked at a lot of data on marriages.

Basically, the next time you get into an argument with someone you're thinking of marrying, note how the two of you resolve it. According to Gottman, if you resolve the argument without causing each other a lot of pain, that's the key factor.
posted by surenoproblem at 5:23 AM on October 3, 2010 [15 favorites]


I am not sure why you think this is under-researched. Do a PsycINFO or Sociological Abstracts search for keyword "marriage" and spend the next five years of your life reading the research. Unfortunately, what you are looking for are tests that you can perform on individual marriage partner prospects to somehow predict future actions. This is not what you will find, as social science research focuses on mean responses to stimuli/interventions/etc. So, research in this vein will report that, on average, subjects in condition X reported emotion Y, performed action Z, what have you, at a statistically significantly higher/lower rate than participants in condition W. However, there is significant variation around the mean, meaning (ha!) that individual participants may or may not closely resemble other participants like them.

However, if you want to get started, you might get started with a classic like Dutton and Aron (1974) in which participants who crossed a fear-arousing bridge before talking to an attractive research assistant were more likely than non-fear-arousing-bridge-crossers to contact the research assistant for a date. The phenomenon is called misattribution of arousal.
posted by proj at 5:26 AM on October 3, 2010 [8 favorites]


Also, John Gottman's work is highly controversial; his "predictive model" was derived from data in which the outcomes are already known. See Andrew Gelman's (immensely respected Columbia statistician) critique of it.
posted by proj at 5:28 AM on October 3, 2010 [4 favorites]


There are tests like FOCCUS give in premarital counseling that help partners identify areas of compatibility and concern. There are some red flag questions, mostly about abuse and addiction, but generally they identify areas of agreement and disagreement in major life issues, such as religion and values, parenting, sexuality, finances, marriage readiness, communication, and problem solving.

Other than the "marriage readiness" section and the abuse-type questions, it isn't so much, "Here's what's right" but rather "Do you AGREE on what's important and how to handle it."
posted by Eyebrows McGee at 5:50 AM on October 3, 2010 [3 favorites]


I think it should be a requirement for couples to go camping or backpacking for a week with just each other before marriage. Even if you don't like camping. *Especially* if you don't like camping. How do you handle boredom? How do you problem solve? Who is the whiney-est about mosqitoe bites?
posted by shortyJBot at 6:01 AM on October 3, 2010 [2 favorites]


Knigel, I really hope this is another 'conversation starter' type question.

You're getting some great answers from proj about attraction research and how to find it.

On the individual level, however--and I understand you're not in the US, so there's probably some culture gapping here, as the US is a very individualistic culture--these are things people have to work out for themselves, and my totally and utterly non-scientific, my-social-framework opinion is that it's less about testing the other person than it is about having good self-insight.

The better you know yourself (and the better your partner knows him or herself), the better you are each able to gauge whether you're a good fit, and how you each contribute to your individual relationship dynamic, and how you negotiate with each other. And even if all of that works out at a particular point in time, nothing is guaranteed. Because life isn't guaranteed, and people change over time. This also goes for any 'test of compatibility.' Let's say that one does exist. It's still not a guarantee, because that's not how people work.
posted by Uniformitarianism Now! at 6:01 AM on October 3, 2010 [2 favorites]


The Relate Institute ihas a widely respected test in the marriage and family field.
posted by WhiteWhale at 6:13 AM on October 3, 2010


I think you mean "test" not "experiment." The only "experiment" you could perform to see if you and you partner are compatible would be to marry them and see what happens. If you're still happy married decades later, you can say 'we must have been pretty compatible.'

Human behavior is not very predictable. The only predictions we can make are probabilistic with greater or lesser degrees of certainty, there are simply too many factors involved. If we have statistical evidence that people with personality trait X are less likely to get along with people with personality trait Y than other people with personality trait X, and you have trait X, all we can say is that, over time, if you marry people with traits X and Y, it's likely that more of your marriages with people with trait X will be successful than those with Y. But you're not doing that - you're marrying one person once. If X-Y incompatibility is a source of conflict in some cases, does that mean it will be for you? Dunno.

Personality tests like Myers-Briggs, which claims to predict people's compatibility with different types of jobs, have been criticized as not very accurate.

If you're concerned about it, you could see a marriage counselor. There's no reason you can't do that before you get married, rather than after, and people do do this. Counseling more art than science, even though counselors are aware of research. And it has to be because counselors are dealing with individual couples, not populations.

(If I was a counselor, I think I'd be wondering about your desire to do tests on your potential spouse to test compatibility, and how your potential spouse reacts to that. What does that say about compatibility? What are your concerns? Why do you want to do that? Of course, there'd be a lot of other questions to ask too. And I'm not a counselor.)
posted by nangar at 7:00 AM on October 3, 2010


which would be the best types of testing that one could do on a potential mate.

Observing their parents would probably be as enlightening as any kind of test you might be able to do in the real world.
posted by K.P. at 7:03 AM on October 3, 2010 [4 favorites]


Travel to a foreign country, drive together for more than four hours, get stuck somewhere (like an emergency room), argue with an intelligent jerk at a party, be really "stuck in the bathroom" sick, move in with each other, any of these things or all of them. The relationships that didn't work out were the ones that we didn't mesh well in crises or lulls, the ones that have are the ones where one carried when the other rested or we both super-charged or we faltered together and were okay with it, but they were also the ones where we maintained our sense of humor and left with good stories.
posted by history is a weapon at 7:07 AM on October 3, 2010 [4 favorites]


Darn, I was going to say what history is a weapon said. Seriously, in past relationships, the ones that flew well and for the longest were those who could pass these "tests without being tests" the best.

The trick is, these can't be contrived, they have to just happen. You can't arrange to be sick as a dog, or lock yourself in a room with your potential partner... but if and when it does happen, I've found that scenarios like these will give you the best idea of what life will be like in the long run with your significant other.
posted by patheral at 7:41 AM on October 3, 2010


To explain what I meant by "not very predictable," here's a statistic I remember from my days as an undergrad psych major interested in group behavior: An overwhelming majority in favor of a particular option (say, four to one in a five-person group) predicts the actual group decision after discussion slightly over 50% of the time (I think about 51%) if there are multiple options available. Sometimes the one person in favor of option B thinks B is better for a reason the others hadn't thought of, and when he explains it the other people change their minds. Sometimes in the course of discussion, option C, which none of them had considered at first, starts looking better. It happens a lot.

Psychologists can learn things about how groups make decisions by conducting experiments, recording the discussions and final outcomes, but they can't predict what decision an individual group will make very well.

When it comes to counseling couples, if something is known to be a potential source of conflict, the question has to be, is it an actual source of conflict for this particular couple, or have they avoided it? If it is, can they compromise so it's not a problem?

I'd be wary of relying on the results of a marriage compatibility test alone, though a marriage counselor might give you one.
posted by nangar at 7:53 AM on October 3, 2010 [1 favorite]


On another note, this is at least the third question (cf., 1, 2) that you've asked looking for psychological "tricks" or shortcuts to reading people. If you are a psychology major, you might want to start looking for these answers in your class readings; if you are not a psychology student, you might want to take some psychology classes. You may soon discover that psychology, as a social science, is not about tricks, reading people, or anything like that. Unfortunately, people are complex, endowed with agency, and react differently in different situations. There are trends that psychology can identify, but you will not find the kinds of answers you have sought in past AskMe questions.
posted by proj at 7:55 AM on October 3, 2010 [7 favorites]


I think your question is way too broad. Take a particular risk and it's probably easier to answer. Risk of spouse dying early, risk of spouse turning into an abuser, risk of spouse pretending to love you for your money, risk of spouse gaining weight, risk of spouse exhibiting trait X, etc. -- various tests can provide some evidence that can shed light on these risks.

I don't think this is a good approach to marriage, with rare exceptions. If you're fantastically wealthy or if you have a history of choosing abusive partners, for example. Otherwise, I think the whole point of marriage is that you face whatever comes together.

I think using divorce statistics to make decisions early on is particularly misguided, since divorce is not something that happens to you but something at least one of you chooses to do. So if you had evidence that people who come from divorced families divorce more (I don't know if this is true) does that mean you shouldn't marry someone who comes from a divorced family? I think that's ludicrous. Statistics about people are useful in the very large aggregate, but not when it comes to individuals (except for some medical decisions, I guess.)

Finally, I think it's clear that intuition provides a lot of information that can not be gleaned from statistics. Even in your car example, you can use statistics to determine likely durability and whatnot, but can they tell you whether you're going to love your car?
posted by callmejay at 7:58 AM on October 3, 2010


I can but share my experience in my 2nd marriage. Was told that going with a woman to a foreign country for a few weeks and not being unable to speak the language of that country would force the two of you to discover how suited you were to each other. We did and passed that test.

Went to visit friends in Maine. We two went in one canoe, the other married couple in a 2nd canoe. We had never cannoed together or even really apart. Canoe paddling of course requires a lot of cooperative paddling. Our hosts told us that how a couple did in canoeing would suggest how compatible they were. We beat our experienced friends by a long distance. Three weeks later, they told us they were getting divorced.

We have been married now 27 years.
posted by Postroad at 8:06 AM on October 3, 2010 [2 favorites]


Response by poster: Proj, I like the the bridge experiment that you posted. That is the kind of answer that I'm looking for. I would like to keep this as much on topic as possible without derailing--thank you~
posted by Knigel at 8:33 AM on October 3, 2010


A friend told me a story about his friend who was about to be married, and before he flew in to Seattle to see his bride, he shaved his head. This was in the 60s, I believe, so it wasn't the fashionable thing it is today. It was a test. He figured if it didn't make any difference to her, she was the one. Apparently it didn't make a difference, and they were married. So you could try making yourself less physically attractive, but it cuts both ways. If your fiancée (assuming you are male) shaved her head or walked with a permanent limp, would it make a difference to you? I don't think too many of us would pass this test.
posted by weapons-grade pandemonium at 8:40 AM on October 3, 2010


the canoe trip....that's a good test.
posted by patnok at 9:27 AM on October 3, 2010


I've said the same thing as the canoe example for tandem bicycles, which I call "argument machines."
posted by Cool Papa Bell at 9:38 AM on October 3, 2010 [2 favorites]


proj isn't derailing. S/he is making an excellent point about how these studies and 'test scenarios' work, Knigel--especially if you like the example s/he provided. You need to have a good understanding of the context and constraints of the field and of the individual study itself--these tests aren't like water displacement or pH tests or testing reflex arcs.

Actually, let me use that last one as an example: even reflex arc tests are subjective and context-dependent, because bodies are weird, and don't always read the book. Sometimes you decide something is hyperreflexia after you test other sites and reinterpret your original 'borderline' decision. Sometimes people appear to be missing a patellar reflex, but they're perfectly normal, which means some variable that you don't know about is messing with the result. And basic monosynaptic reflex arcs are about the most predictable and clearly understood tests of anything involving the human CNS. They are as simple as it gets, and honestly, they're not that simple.

So when you get into things that involve higher consciousness and decision-making and personality and predicting social behavior....you are light-years beyond the not-actually-simple knee-jerk reflex into the realm of conditionals and context-dependence and 87000 other complicating factors. I don't mean that in a bad 'psychology-sucks' way, either. I mean it's just the nature of the problems psychology researches.

Look, I took a series of standardized neuropsychological tests this past week. And they're really useful and I support them, but at the same time they're founded on a lot of assumptions and pretty-good-as-far-as-we-know estimates across populations. They have to be--nature of the beast here. And these tests didn't involve anything about long-term potential with my partner. That's a few levels of complexity and synthesis above what those tests were measuring, which is a set of traits that are already pretty freaking complex, and there are good reasons you have to be trained to administer the tests.
posted by Uniformitarianism Now! at 9:51 AM on October 3, 2010 [3 favorites]


Mod note: few comments removed, ease off the snark please? thanks
posted by jessamyn (staff) at 10:41 AM on October 3, 2010


Does anyone know if there are reliable statistics about marriages that start in the mid-30s-to-40s compared to marriages that start earlier? Most of my friends who married "late" are in successful marriages. The ones who married early are mostly divorced.

Also, the ones who married really good friends are doing well. The ones who married people they had "chemistry" with but not intimate friendship are not doing so well.

I know neither of these guesses are Earth-shattering, but Marry Late and Marry Your Best Friend seem to be the two best tricks for attaining a happy marriage. So I'm wondering if the data supports this.
posted by grumblebee at 11:32 AM on October 3, 2010


Knigel, I'm confused as to how the bridge experiment answers your question -- it concerns subjects meeting someone for the first time. Are you interested in arranged marriages?
posted by yohko at 11:58 AM on October 3, 2010


Response by poster: Grumblebee, actually the bridge experiment isn't so useful for me; however, I wanted to encourage people to post that kind of information instead of bringing in irrelevant threads and presumptions about my understanding of psychology.
posted by Knigel at 2:17 PM on October 3, 2010


Response by poster: Sorry, Yohko, not Grumblebee
posted by Knigel at 2:18 PM on October 3, 2010


Response by poster: Uniformitarianism Now!, yes, Psychology does have its flaws; however, this thread isn't to discuss the obvious problems of psychology as a science, I am looking for experiments and tests that could be done on potential marriage partners.

Even without 100% accuracy, various psychological tests and experiments could be done that may certainly do a better job than simply following one's love-stricken intuition. We continue to use psychology in our society even though it isn't always perfect; therefore, it seems rational to at least consider its uses for potential marriage partners.
posted by Knigel at 2:41 PM on October 3, 2010


Science is made up of falsifiable theories. Divorce is the falsification of "I do." Look at the statistics for divorce to get scientific data on marriage. Age ranges, custody, settlements, court costs, sex, division of property, location, socio - economic status, previous marriages, etc. Patterns do emerge.
posted by eccnineten at 3:03 PM on October 3, 2010


Psychology does have its flaws; however, this thread isn't to discuss the obvious problems of psychology as a science, I am looking for experiments and tests that could be done on potential marriage partners.

One company that agrees with you is eharmony. They offer products not only for people looking for marriage partners but also for people in committed relationships (I am pretty sure about this). Why don't you check out their website and methodology?

Eharmony boasts of the science and psychology behind questionnaires and other testing. A lot of it is proprietary but they make enough of it public to attract customers.

I agree with the advice you've gotten here about the problems with what you're trying to do, but Eharmony sells the customized, quantifiable thing you're after, and you can access it from Korea.
posted by vincele at 3:20 PM on October 3, 2010


Response by poster: To be honest, while I did mention personality tests, I am looking more for advice on behavioural tests and laboratory experiments such as raising the room temperature while the potential hubby is performing tasks. These types of tests could bring out temperament and other more stable personality traits.
posted by Knigel at 4:08 PM on October 3, 2010


Some people are good test-takers, others not so much. Keep that in mind.
posted by yesster at 8:00 PM on October 3, 2010


Knigel, the kind of experiments you're asking for are tests that aren't pencil and paper tests, that is, performance tests. Some of answers here are examples of that - go camping, take a foreign trip, the canoe test. The first two test how well you do as couple when you're by yourselves in kind of stressful conditions. The canoe test uses paddling as a test of cooperation.

The problem is you want procedures used in psychological experiments, and the kind of test you want isn't what psychologists would call an experiment. This online textbook gives a fairly concise explanation. Marital compatibility doesn't lend itself to experimentation. To do an experiment you have to hypothesize that one thing (Y) affects something else (Z), and be able to control Y. You can't give a person a personality trait to see how that affects their relationship. So you're back to surveys and personality tests.

So, are there any performance tests for marital compatibility developed by psychologists and derived from procedures used in psychological experiments? As far as I know, there aren't any.

This isn't an attempt to derail. I'm just trying explain why there isn't an answer to your question the way you posed it.
posted by nangar at 8:11 PM on October 3, 2010 [1 favorite]


Response by poster: Nanger, I do mean experiments; however, I am not so focused on compatibility exactly, only things that would bring out some negative or positive issues.

For example, my independent variable could be decreases of heat and the dependent variable could be signs of aggression perhaps operationally defined as swearing or rude comments towards another participant (perhaps even children) working with them on the task. As there is only one potential candidate, we can't use a control group and experimental group; however, we can do a within-subject experiment.

If I had two prospects, and one of them uses breathing exercises to control their anger while the other one flips over the table, we could determine that the experiment was successful in adding one point for one candidate. Would the aggressive one usually act like that? Would they ever act like that again? Maybe, maybe not; however, I think that would be a bit of evidence to get the gist of their temperament. If my prospect smacked me during a stressful lab experiment, I could determine that they are not the one for me. It's objective data.

This is one type of experiment that can be done. There are many kinds of experiments that can be done to bring out potential future problems. We can even throw them into a Skinner Box to see if they are prone to gambling.

There are also many tests that aren't strictly experimental. Some examples would be variations of the chocolate test--do they eat the chocolate now or wait for two? Are the impulsive? We can see how they play some games such as the prisoner's dilemma.

With both experiments and tests we can find information about how they handle money, how they treat children under pressure, if they become aggressive, if they become humorous through a difficult task, if they give up easily, if they are stubborn, if they are prone to taking charge or letting someone else take charge, if they are a high/low-monitor, if they think more creatively about a problem or follow rules. Do they follow a crowd when they know they are wrong about something or stand out? There are so many things to test.

What I'm trying to get at is that there are many ways we can use psychological techniques, both tests and experiments, to get more objective data about someone that has very serious potential risks. I fully understand that each kind of study, test, survey, case study, or experiment that we use has its own flaw; however, the scientific method does serve the purpose of allowing us to not fly completely blind. Our intuition, even though it has major strengths, does often fail us when it comes to relationships.

We might not always gain the best data; however, it seems to me that there are many ways to get objective data about potential marriage partners that we wouldn't have without running some scientific studies.

If it helps, pretend I'm writing a science fiction novel and am looking for some cool experiments and tests to run in some dystopia.
posted by Knigel at 9:10 PM on October 3, 2010


Response by poster: Since we are already heading in this direction.

Proj, I do not know why you posted those other threads of which were about entertaining people and not so much about reading people. You said it perfectly that psychology has certain trends. These trends can be used for various reasons and they are often more successful than trusting "common sense". They can help people get over phobias, be fun to trick people with optical illusions, or interesting when exploiting mental blind spots or cognitive bias.

I've never seen your definition of psychology as a social science before. From my understanding psychology has many functions and is about many things. Psychology is often very much "about tricks, reading people, or anything like that". That's not all it is and people are complex; however, while we are still gaining objective psychological data; tricks and attempts to read people are a part of psychology.

To say otherwise is like saying chemistry isn't about blowing things up for fun. Some people love doing that; therefore, it is very much about that for them. Some people can delve into cognitive science and help people through CBT while others can study optical illusions because they are just so trippy. I personally find many aspects of psychology engaging and I often get great answers from making posts here.
posted by Knigel at 9:45 PM on October 3, 2010


... my independent variable could be decreases of heat and the dependent variable could be signs of aggression perhaps operationally defined as swearing or rude comments towards another participant ...

What you're looking for, though, is how a particular individual responds to that condition. An experiment would ask if decreases of heat affect aggression. You want to know how much a particular person is affected by it, and presumably compare them to an already established baseline to see if they're more or less aggressive than most people in response to those circumstances. In other words, it would be a performance test, not an experiment (not in the strict sense).

You seem to be interested in tests you could perform on a potential partner to see if they meet certain criteria you have in mind, not in tests you could also take, or you and your partner could take as a couple. I still find that a bit odd.

If my prospect smacked me during a stressful lab experiment, I could determine that they are not the one for me. It's objective data.

Not more so than it would be if they smacked you in a stressful situation if real life. The point of lab experiments is that psychologists can observe what happens and control some of the variables in the situation. For participants, being in a psychological experiment is part of real life. What you do in a psychological experiment isn't more predictive of your future behavior than what you do elsewhere, just because it's in a psychological experiment. Being observed doesn't make it more predictive. (If you're the observer in question, it doesn't matter, and real life situations might be better.)

(I'll have to add, psych researchers wouldn't use slapping as a measure of aggression in an experiment. There are ways to get people to think they're hurting someone in an experiment without letting them actually do any harm.)

In a dystopian world where ethics didn't matter, it would be possible to imagine performance tests for things like abusiveness, jealousy or loyalty that could applied to couples or members of couples. There would be other problems with such tests beyond just ethics. Performing them could permanently alter the relationship between the couple, even if they "passed" it.
posted by nangar at 11:23 PM on October 3, 2010 [1 favorite]


It sounds like you are not so much looking for compatibility between two people as to find hidden flaws in someone else.

You mention tests for new employees -- maybe some of the personality testing that some places have prospective hires go through would be the sort of thing you are after. Has it simply not occurred to you that these types of personality tests could be taken by people who are not job seekers? Is it that these types of assessments are usually pencil and paper or online, rather than in a lab room with the heat being turned up?

You seem to be only interested in testing the characteristics of one or two individuals to see if anything comes up that you don't feel you would like -- if someone was going to study marriage compatibility, it seems to me that they would need to look at both people. If you want to see if the person has anger issues, what difference does it make if you call the subject "prospective spouse #1", or "prospective head of widget division #1" -- they would have the same personality traits persisting across time either way.

On the other hand, finding a prospect willing to go through these lab experiments may function as an ersatz compatibility test between two people -- prospects who feel that they are incompatible with someone who has been unable to notice their personality traits by observation over a period of time during their relationship thus far may be more likely to show anger during the lab test, possibly as they announce a lack of further interest in continuing to be "prospective husband #12".
posted by yohko at 2:23 PM on October 4, 2010


Seriously, there are a ton of physical and chemical predictors that determine long-term compatibility between spouses (many are hormonal... oxytocin production and brain activity are superior indicators of "lasting love" and remain strong in partners of 30-plus years of marriage, even into their 70s, even while looking at a PHOTO of the partner and nothing else).

Study 1

Three generations of data showing what men and women rank as important in a husband/wife candidate, 18 factors, highest to lowest

Dr. Helen Fisher's research - check out especially the brain scan data that's been in National Geographic and stuff

I think your wording is confusing the larger issue, which is this: THERE'S A CRAP-TON OF DATA ABOUT THIS.

Also: there are physical traits that people have which indicate behavioral tendencies, and to test a partner on things like attraction level and whether they're lying or flirting with you in vain, just look up profiling techniques from retired polygraph testers, body language experts and FBI interrogators.

Let me know if you want more, I could do this all day.
posted by Unicorn on the cob at 9:11 PM on October 6, 2010 [1 favorite]




« Older How to get smooth hair   |   Tuned in, turned on, but dropped out. Newer »
This thread is closed to new comments.