91 comments

[ 3.1 ms ] story [ 166 ms ] thread
Having a hard time finding causation here
Relationship movies are marriage propaganda.
Just a few really strong movies and guided discussion after that can definitely completely change ones perspective towards how to approach their partner; this study seems to affirm the same.
You shouldn't, interventional studies yield causal evidence, by definition
The causation is the same as for common pre-marital counseling methods, having an improved shared understanding of relationship expectations (including improved understanding of partner expectations where they differ from yours) before running into clashes of expectations in marriage.

This is done early marriage, and a slightly novel specific format, but it's not really all that different from pre-marital counseling methods which have been known to provide good results for decades, which have often included, which have often included pretty similar time allotment to the mechanism here, and a pattern of spaced sessions with presentation consisting largely of scenarios (fictional and/or from other couples) followed by time for the couple to discuss. So I'm not at all surprised that this is both effective and as effective as intense but narrow focus on one or anothee tool for coping with stressful situations.

(comment deleted)
Drops divorce rates on what timeline? Reminds of Intervention, where all addicts relapse within 6 months.

I’ll never understand how divorce happens when two people have kids. Anything before that is just a dressed up break up.

The article starts off with, Discussing five movies about relationships over a month could cut the three-year divorce rate for newlyweds in half...

So it reduces divorce rates in half.

I would be curious at how large the sample size was though.

> So it reduces divorce rates in half.

Over three years, not over the life of the marriage, but, while the average life of a first marriage that ends in divorce is around 8 years, there's a lot ending in the first couple, so having the three-year divorce rate is a big deal.

That and to do a study, you have to define your success criteria. There is a big difference between taking 3 years to get a result and taking 8.

Even though 8 would be more meaningful, I completely support looking at 3 instead and hoping that it is meaningful.

For the people in the relationship, taking 3 years to get to a [relationship] result may be better than 8 as well.

If I’m going to end up divorced in all cases, I’d much prefer it only take 3 years (though others might prefer 8).

So, if this counseling activity moves divorces out, it could be viewed as worse by some.

This makes sense.

There is other research indicating that intellectual knowledge does not translate into behavior. But direct or imagined experience does.

So for example someone who can give all the right answers on a psychology test still acts the same under pressure. But people who have had to go through and reflect on simulated experiences will have that change actual behavior.

Therefore watching and discussing relationship movies is in line with what is known to actually change behavior.

The two books that made this point most effectively to me were Enlightenment Now and Thinking Fast, Thinking Slow. The first in the context of how much the rise of fiction changed the amount of empathy that people in general had. And the second explains it in terms of how intellectual knowledge only helps with the "thinking slow" parts of our brains, but you have to affect the "thinking fast" parts to change actual behavior.

Or...

Fiction drives home unrealistic standards or extremes.

I found the data to be dubious ofcourse. Just like we can extrapolate all sorts of nonsense from data - movies aabout relationships might not mean anything at all.

Considering that life is 90% boring routine and fiction does not show or truly represent that boringness in order to sell the story. They need drama. Thats why writers wont write about a perfectly normal day with a couple talking about tiles for an hour dialog near the kitchen sink.

The movies served as means to make couples talk about their own relationsship. Check the questions from the PDF that I linked elsewhere here to get a better idea.
> The movies served as means to make couples talk about their own relationship.

Sure. The next one should be titled Taylor Swift lyrics decrease divorce rate by 45% but Nicki Minaj lyrics don't.

I'm sure the methodology will hold up equally well.

Marriage ad the strongest legally recognised relationship outside blood relations - needs interventions so that couples can talk. Is this an approximation of the problem statement?

If it is, either no one really knows the problem or it is known but won't be touched with a 100ft pole since doing so will lead down the rabbit hole of disillusionment.

I dont understand what are you trying to say here.
The article, methodology and data are dubious.
| Fiction drives home unrealistic standards or extremes.

Strongly depends on the fiction.

500 Days of Summer or Eternal Sunshine of the Spotless Mind, for instance.

These movies require you to use meta-cognition to reflect on why the main characters are doing what they're doing.

They are a psychological antidote to the the hero's quest of unrequited love.

If anything propagates the fiction of unrealistic standards it would be Disney movies.

> intellectual knowledge does not translate into behavior

Wow I'd love to see some links supporting that, it feels hard to believe!

Basically every gambler is an excellent example of that principle, though I have also not seen the research.
Treatment for problem gamblers uses mathematics to prove they cannot beat system

https://www.abc.net.au/news/2014-11-30/new-approach-to-treat...

Okay now that we got the intellectual argument out of the way.

Let's talk about the actual experience of gamblers: You're not playing against the house. You're playing against the other gamblers. This is highly prominent in Poker where you try to beat other players.

Some gamblers told me about their weird theories about which machine will give them the fattest payouts because they analyzed which machines have been used the most and how to minimize the opportunity cost of their current wager by taking advantage of free services provided by the casino. They clearly aren't in it to stick it to the house. They are trying to outcompete other gamblers with their strategies even with a slot machine.

It makes intuitive sense to me. Studying a thing and practicing a thing are different. We practice fire drills for a reason.

And you may have intellectual understanding of what plays to run on the football field but it doesn't make you a good quarterback.

If you haven't read 'Thinking Fast, Thinking Slow', give it a look. It is one of the most influential books I've ever read.

That will explain it. I could reach out for my copy and look up citations, but probably searching on the author Daniel Kahneman would get you what you need.

You should say no to a purchase.

Wife and kids will all melt down if you say no. The last meltdown ruined a weekend.

The purchase isn’t that big a deal. But it’ll make other important bills late. But if you say yes, it’ll encourage the behavior. But you will also have a few days of peace.

Knowing the right thing to do, and having the energy and will power to follow through are very different things.

I don't think this is actually an example where you really know what the right thing to do is.

A) You can do what you think is right

B) You can do what other person thinks is right

It's not clear that either A or B is the correct choice.

It is when B is financially impossible and/or will lead to huge problems later and/or is only motivated by short-term desire which won't be satisfied by a passing whim.

It's not a gendered problem, although of course it looks like it when the other half of the couple is demanding The Thing and you're the one saying "no".

Or vice versa.

Some actions are strategic, and some are whimsical.

Whimsy is welcome and enjoyable when there are no strategic consequences, but not such a good idea when the consequences are obvious, predictable, and negative.

Just because you are the one who said no does not mean the other one is whimsy or buying stupid things.

It really depends on whose non essential wishes are getting more funds in familly overall, whether there would be a strategic problem and so on.

But if your partner is annoyed so much over how money are spend, something is very wrong. Because regardless of who is right or whether split is fair, you are building long term ressentment problem. The attitude of "I get to have last word them partner not liking it is childish meltdown" destroys relationships.

I recommended 2 books because I know it from books, not random blog posts.

But both books are classics.

Quick self test: Do you eat healthy food, floss your teeth, get enough sleep, care about your relationships, and exercise somewhat? This is boring simple common sense advice everybody knows but I have not met anybody (myself included) who achieves it.
You don't know anyone who mostly eats healthy food, usually flosses, usually gets enough sleep, puts a reasonable amount of effort into their relationships and exercises somewhat? Or you don't know anyone who does all those things perfectly?
For all the confused commenters, heres a hypothesis:

Talking about relationship movies gives a couple a platform to discuss issues that affect their current relationship, with much lower stakes than in the middle of a conflict. This doesn't have to effect change so much as it enables the partners to understand each other better, and accept the other more fully when conflicts arise. This can also enable the pair to adapt to each other better.

My girlfriend and I do this all the time, coincidentally, and i find it to be a great way to learn about her and to discuss how we might react in situations that would be difficult to resolve otherwise (like all the ones on the tele).

TLDR: It cuts the tension.

Most hollywood movies also present relationships in a positive light, and people are memetic creatures.
I have not managed to get access to the paper yet but I wonder if they have considered "treating" only one of the partners to see how much that helps. Is the "treated" partner able to provide almost all of the advantage on their own? Or is the other partner the "blocker" then, making the effort completely futile? Does it work to go through the process without being in a partnership at that time or does it require immediate practice to have a lasting effect on the behaviour? There are so many interesting follow-up questions...

Edit: after reading the questions about the movies that the couple were asked to answer, I doubt that my questions actually make that much sense.

Deep Blue Something was onto something.
Unfortunately, in their case, it seems to have been too little, too late.
The link to the list of movies and questions is broken in the article but it can be found here instead:

https://www.rochester.edu/news/divorce-rate-cut-in-half-for-...

Edit: it's also in the right sidebar but I did not notice it until I had seen the PDF already (the graphics is the same as the frontpage of the PDF). Funny how being on the interwebz long enough trains one to ignore everything but the main content.

Can’t access the paper but deducing the following numbers from the abstract :

Participants per group - 43

Number divorced in control - 11

Number divorced in all three treatments - 5.

It sounds iffy but when I think about it, it probably will pass most statistical tests of significance. So this is surprising for sure.

I find divorce statistics hard to fathom. A rate that compares numbers of marriages and divorces in a single year does not really seem to speak for long term prospects, especially due to population changes over that time.

For example, are there any reliable statistics on the percent of people who married in 1970, who did not divorce to the present day? And for 1980, 1990, 2000 etc.

There’s a statistic that the average length of the American marriage is 8 years.

Yes you’re right, the overlap numbers are incredibly confusing. But the 8 year number is a pretty good indicator that people don’t remain married for long anyway.

https://www.census.gov/prod/2005pubs/p70-97.pdf

Under a standard bell curve, what percentile would a 20 year marriage be then?
If it was normally distributed, it would be 100% - the percentile of a -12 year marriage. So it's surely very much not a normal distribution.
Normal distributions are used all the time to model strictly positive phenomena. https://math.stackexchange.com/a/1335872
That's alright when the negative values are way off the end of the tail, but a 20 year marriage is way more common than a 3.4 meter tall woman (same distance from the mean as a 0 meter tall woman).
That the marriage distribution by age and the height distribution of women might have different decay rates is very unsurprising. That does not demonstrate that using a normal distribution to model either one is inadequate.

In order to actually demonstrate that a Gaussian curve is not a good fit for the distribution of marriage lengths, you’d need to show that the falloff on both sides of the “mean” is asymmetric, or that there are multiple peaks, or that the falloff does not trend exponentially, etc. Even a truncated normal distribution might fit the distribution of marriage lengths adequately.

Figure 5 in the more recent data shows the CDF of the distribution of marriage lengths: https://www.census.gov/prod/2011pubs/p70-125.pdf

You can compare that CDF to plots of the CDF of a normal distribution: https://en.wikipedia.org/wiki/Cumulative_distribution_functi...

If you can translate & scale any lines from either graph to roughly match, then you can use a normal distribution to roughly model marriage lengths.

BTW, we are both being pedantic. The grandparent didn’t claim it was a bell curve, they asked what the standard deviation would be assuming it was normally distributed. That’s something we can calculate, regardless of how normally distributed the data actually is. It’s a strawman to say the answer won’t be exact because the question assumed it won’t be exact.

Also, I do not think marriages are exactly normally distributed. We could calculate some stats about how close they are or how far away they are. There is a single lump shape to the distribution, but there’s also some asymmetry, and the distribution is truncated on the left. My point is just that the two reasons you’ve brought so far aren’t showing that marriage rates are not a bell curve, you need better reasons that are more specific to the shape of the data in order to make that conclusion.

I don't see any reason to think it follows a bell curve; it has to be skewed way to the right, as it can't be negative. The question you're asking is more-or-less answered in eric4smith's link, on page 6. The answer depends on the year of marriage, and the data is from 2001, so for 20 years of marriage, it can only go up to 1980. The percentage of men whose marriage had lasted at least 20 or 25 years, by year of marriage, was:

                20yrs 25yrs
  1955 to 1959  76.2  72.3
  1960 to 1964  66.1  62.3
  1965 to 1969  62.1  58.0
  1970 to 1974  55.8  52.9
  1975 to 1979  58.4
> There’s a statistic that the average length of the American marriage is 8 years.

That statistic is not in the census report to which you linked. Instead, the report says on its first page that “First marriages that end in divorce last about 8 years, on averaged.“

Please make sure to read the full paper before jumping to conclusions! This was not a fully randomized study!

The main outcome of the study was that couples that participated in ANY treatment had similar outcomes. The benefit was compared to a "No treatment" group.

But the article fails to mention that this "no treatment" group was not randomly selected!! The "no treatment" group was made up of couples that refused treatment, or failed to participate in treatments because of scheduling constraints.

So a less misleading headline might be: Couples that refuse relationship counselling have higher divorce rates. But that would surprise noone.

It's getting really tired how unreliably scientific results are reported in the mainstream media.

Is anyone building a (crowdsourced) fact checker to validate these kinds of claims that I could install as a browser extension?

The worst part is this isn't even a mainstream media source. This is coming from a university!
(comment deleted)
They may have even stronger incentive to juice up a press release.
It is soft science. Almost all results are unreproducable BS
Right, I would like a study on how discussing relationship movies increases the marriage rate for couples.
> Couples that refuse relationship counselling have higher divorce rates. But that would surprise no one.

It would surprise some; it's perfectly possible to take the view that counselling causes divorces.

(comment deleted)
Even less adventurous would be to consider that counselling has no effect. But in any case, the study can only provide information regarding counselling with respect to couples that refuse it, not regarding counselling in general.
Scott Alexander wrote he has no clue if counseling works:

> I tell patients to have a consistent weekly meeting with their spouse, maybe a Saturday date night. During the meeting, each of them has to say one nice thing about the other, and then one problem that they want to work on. The other person thanks them for the nice thing and then they brainstorm a solution to the problem. If they both agree to the solution, they have to stick with it at least until the next relationship meeting. A lot of my patients have said this really helps them, have continued the meetings long after the immediate crisis passes, and probably expect the technique was invented by some clever person like John Gottman. I will never tell them that actually I picked it up from a different ex-girlfriend, not even the one who writes relationship advice columns.

> In conclusion, I have no idea what makes marriages work, and I am not convinced John Gottman does either.

https://slatestarcodex.com/2020/02/27/book-review-the-seven-...

I find the speculation on what the patients think reaching. I also find the origin in terms of which girlfriend of his first suffested it completely irrelevat.

It is standard relationship advice that flies around in various forms. I have seen multiple variants of it - the two componentams are regular dates basically and system which creates safe regular times for dealing with issues outside of conflict situation.

I mean honestly, the above does not give much trust to him.

I think the point is that he learned (and presumably implemented) the technique in a relationship that failed and yet he recommends it without sharing that context.

“I recommend this construction technique.” “Sounds great. By the way where’d you learn it?” “I learned it building my last house.” “Can I go see it?” “No, it collapsed.”

Houses are meant to stay upright, but relationships are meant to end. Consider all your relationships (romantic or platonic) from middle school, high school, uni, first job, moving to a new city, cofounder etc. the expected, healthy, and natural conclusion for 90% of your lifetime relationships is to end to make room for the next ones. You learn from each one and try to apply from the best to the rest. Sometimes, very few times in life, an end is a failure - but the baseline is that it’s natural and desirable. Unlike your house collapsing.
Relationships certainly change but it’s depressing to assume they will all end. Actually most of my relationships have not ended, except mainly when the other party has died.
That would assume that the above technique is silver bullet that is meant to solve all possible relationship problems.

Such a thing just can't exist. If that is the standard, then you would give up on giving to fix anything in relationship. Since there is a couple that whose issue won't be solved by that, why bother.

Especially with new relationships, you two are basically figuring out whether the couple have compatible value systems, expectations, whether one of them has destructive tendencies it is safer and better to nope out of. It is ok to break up if with the "I realized this is not for me, I think I will be better off with someone else".

If all problems are emotional, then just the very act of a shared commitment, reccomended by an "expert" (e.g. to having "relationship meetings") may be enough to solve them, regardless of content.
(comment deleted)
(comment deleted)
I think it is perfectly reasonable to think that counselling correlates with divorce. People normally only go through counselling when things go wrong, and divorces also happen when things go wrong. I guess people are more likely to die after seeing a doctor for the same reason.

Refusing counselling against the advise of a specialist is another metric. And while I think that indeed, it correlates with higher divorce rates, it is not obvious at all and it took a few studies to get that result.

Maybe I chose the wrong word. Writing good summaries is hard!

What they describe in the paper (they call it a "treatment") is more like a training for newlyweds.

Maybe I should have phrased it like this:

Couples who decline to take part in free relationship training have higher divorce rates.

Or, more positively:

Couples who participate in free relationship training have lower divorce rates.

(But that implies causality, so it is again somewhat misleading)

> This was not a fully randomized study!

Which makes it practically useless since most observational studies are poorly controlled.

You could also say “discussing movies as effective as therapy” which is quite an interesting find.
Yes, and that is the finding of the actual paper. The kicker is that the movie method takes less time overall and less therapist involvement than the other two techniques tested. I wish people could get past whining about how this isn’t a randomized controlled study and actually look at the findings on their merits.
HN should require (non-randomized) in the title for this kind of thing, the same way it requires (2014) or (pdf).
This is the kind of paper where approaching it with a kind of Bayesian orientation is really helpful. As in: what's your prior on movies cutting the divorce in half? It's almost certainly very small. So either this paper has to have amazingly compelling evidence, or something is wrong.
> So a less misleading headline might be: Couples that refuse relationship counselling have higher divorce rates. But that would surprise noone.

Relationship counseling is delegating responsibility for your relationship to a third party. Or, in most cases, transferring blame for your psychological shortcomings to your spouse via a faux neutral third party.

(I don't doubt that there are good counselor, but they are probably even rarer than good psychologists.)

This makes me wonder about the role of film as a “machine for empathy” as described by Ebert. If movies could prime people for relationship chats by seeding empathetic ideas.

This led me to wonder if there’s any relationship between the media and ability to elicit empathy - if couples discussed relationship books would they also see a lower rate of divorce? So far, I can only find one paper on this (https://doi.org/10.1075/ssol.19003.tur) but I wonder how it fits into the wider literature?

What is known about the ability of different media to generate empathy? The paper above suggests books are better than films. What about immersive games?

I think the fact that they are both willing to watch some of those films must improve chances of staying together
Here’s the paper: https://pdfs.semanticscholar.org/53e2/70a3f29cf556849362130e...

Unfortunately on closer examination the results look pretty meaningless to me. The study involves rather small numbers (e.g., 4/33 vs 10/44 divorces), but an even bigger issue is that the “control” arm in this study wasn’t selected randomly, but rather consisted of couples “declining active treatment” or who “could not be scheduled for active treatment.”

So a better summary might be “Divorce rate cut in half for coupling willing and able to take part in a relationship building program, of any sort.”

If these psychologists had only been a bit less greedy about wanting to compare three different treatment arms rather than a real control arm, they could have set up a study that could actually have taught us something...

174 is a pretty small sample size. This was several years ago, anyone know if there was any follow up?
I can rarely remember seeing a title that made me so quickly and confidently say "Will Not Replicate". I mean, I say that often, but this made me say it louder.
Tip: don't marry in the first place.
Here I was complaining about "big sugar" or "big tobacco" influencing scientific studies, when "big romantic comedy" was behind the wheel this whole time.
This reminds me of an episode on Ozark where the main character secretly pays the psychiatrist whom he regularly visits together with the wife to be on his side.

Didn't take long for the equally cunning wife to successfully attempt to bribe the morally bankrupt psychiatrist.

(comment deleted)
I’m pretty skeptical of feel-good-10-o’clock-news-ready studies like this.

Don’t get me wrong, it could certainly be true but I come in from a place of skepticism.

Maybe a person who has the resilience to tolerate discussing a relationship movie, is also likely to tolerate a bad marriage.