The first time around they got the result they were expecting so they didn't go back and check their code. They probably would have if it initially stated the opposite.
“Once is happenstance. Twice is coincidence. Three times is enemy action”
― Ian Fleming, Goldfinger
P-hacking is so endemic and so essential to modern academia, especially in the 'softer' sciences, that the whole thing is beyond saving. It's so bad that anyone who understands statistics well enough to know that most of their peers works are bogus are much more likely to be shunned than embraced because of this knowledge which is an incredible disincentive to acquiring it.
Edit: as I can’t reply to the now dead account, the first quote explains why academia cannot keep claiming innocence on such matters. The rest of my comment was my explanation as to why nothing is being done about it. Sure, we can ask them to do better but they should already know that.
It's more like, you should be very skeptical of any shocking results that you find reported on widely in social media and the press.
Errors will always happen; they are more likely to result in shocking, surprising results; and the press actively publicizes those results over others. The original paper got widespread press coverage, and its retraction got... Much less. You end up much dumber if you outsource your thinking to the Washington Post or Facebook.
>We see this in a very specific case, which is in the onset of heart problems. So basically its a more nuanced finding. The finding is not quite as strong.
The one divorce I can think of in my extended family came from the onset of heart problems. Properly speaking, it was the abortion she had that caused the divorce, not the fact that she would have died otherwise.
The main concern I see with "technologic" wombs is who is then responsible for the upbringing and care of the fetus until it because viable to live on its own? If someone gives up a fetus but is still force to be financially responsible for it, more harmful and destructive forms abortion will be, by necessity, be used.
Who is responsible for incubator babies until they are viable to leave the incubator? That's already a limited sort of artificial womb. The state picks up the bill I suppose, if everybody else decides to bail.
My use of “unburdened” includes financial in all variations, even sealing the ability of the subsequent child to find the parents.
It’s pretty clear that it will be the state who assumes receivership of the fetus and subsequent human with full constitutional rights.
It’s also a pretty easy legislative problem to say the state is transferring unwanted fetuses, despite the technology for its viability not currently existing. The outcome shifts the burden of guilt - or lack thereof - away from the parent(s), over to an immune state regardless of that state’s laws on abortion. And it shifts priorities and funding measures on creating the technology to ensure a fetus’ viability, whether that turns out to be a fool's errand or not.
This will easily bridge consensus between “camps” as the discomfort over citizens killing fetus turns into a market choice instead of a legislative debate. What will the market primarily choose to end a pregnancy
Hm. Male with heart problems here. I am in a happy relationship now but dating has been hard in the past due to not being able do do certain things normally expected of men.
~43% of first marriages fail, ~60% of second marriages, ~73% third marriages. Lots of people are emotionally unhealthy and barely equipped to adult, let alone have the grit, fortitude, and luck to make a partnership work for life. It is the hardest job you’ll ever have. The divorce rate is understandable considering the partnership expectations outcome mismatch in the aggregate.
Suppose he really wanted kids and learned she couldn't have any. It wouldn't exactly be "until death do we part", but I think much of western society has already abandoned that standard anyway
My goal was a lifelong marriage. My mentality constantly evolved toward that goal.
However, my judgment was insufficient and I didn't properly recognize that sustaining the marriage was harming our kids.
In an earlier time, the harm would have been much less. Kids (my generation & earlier) once had direct, relationship-level access to a number of caring, invested adults - related and not.
My kids had the two of us 24/7 and that's about it. My wife and I had 10x the parenting demands (time) put upon us that my parents, grandparents-> adults through human history had.
> Karraker and her co-author did the analysis again, and found the results stand only when wives develop heart problems, not other illnesses.
This is p-hacking after-the-fact, right? Seems like the classic example: The broader hypothesis doesn’t hold, but if you look for ways to slice and dice the data you’re likely to find a (spurious?) correlation eventually.
It looks like the authors didn't properly handle missing values in Stata, leading to marking people with missing health information being marked as being "severely ill" instead of being excluded from the analysis.
Not OP, but to me it sounds line p-hacking aka bad science as well: If you slice a dataset en enough subsamples you will very likely find random correlations. That’s the nature of these kinds of analyses and we should be sceptical of conclusions that are based on suce analyses.
I think you and p51-remorse are discussing different parts of the article. They're saying the updated analysis is suspect because of the risk of false discoveries. I believe that's probably true in the usual way--if we study 20 subgroups with no actual effect, then we expect one to show an effect with p < 0.05. There's no mention of preregistration or anything like a Bonferroni correction to manage that risk.
You're saying the original analysis was wrong due to a coding error. I believe that's also true, but that's not what they were discussing. The variable names are inscrutable, but the article text also seems to imply that line (mis)codes divorce, not severe illness:
> People who left the study were actually miscoded as getting divorced.
So they actually found a correlation between severe illness and leaving the study. That's perhaps intuitive, if those people were too busy managing their illness to respond.
If you have no statistical significance in any other type of illness but one, the chances of the data for that one type of illness being not balanced are high. They should do a follow up study to verify their findings on heart conditions with a new set of data.
Although you're right that there's likely no relationship between wives developing heart conditions and subsequent divorce, there's not enough information from the article to know whether there's anything statistically meaningful about heart conditions specifically. It seems more likely that it's just statistical noise. I read that section and got the impression that the relationship is interesting but it doesn't necessarily mean anything, rather than implying that the original study still has some validity.
I wondered that but looking at both papers, it seems much less likely. For one thing, they looked at the same illnesses in both the original paper and the new paper, so it’s not like they did a bunch of new analysis in order to find something that’s significantly.
For another, the study doesn’t look at that many diseases, just four: cancer, heart disease, lung disease, and stroke. It’s possible the original study was doing some p-hacking in order to reach significance but these are pretty major categories of disease so they seem defensible. It would not surprise me if these are the only diseases in the original dataset.
Finally, the results are significant at the .01 level. Combined with the number of diseases, this is not anything as blatant as the classic XKCD green jellybean comic, although more subtle p-hacking could be at play.
> What we find in the corrected analysis is we still see evidence that when wives become sick marriages are at an elevated risk of divorce, whereas we don’t see any relationship between divorce and husbands’ illness. We see this in a very specific case, which is in the onset of heart problems. So basically its a more nuanced finding. The finding is not quite as strong.
I take issue with this framing. Starting with saying that when women get sick, risk of divorce elevates, immediately followed by saying this happens only in one type of sickness is still very misleading. The entire first sentence is misrepresenting the actual findings.
It's possible that for experienced researchers, "heart problems" is a more universal category than we're treating it here. AKA a rough analog for "life-threatening"? I'm under the impression basically all deaths due to senescence (old age) occur from heart failures (potentially in turn caused by other things), so "you're cancer has advanced so far that it now is threatening your heart function" seems like something that might happen frequently. Thus, they're defending the smaller "husbands leave their worryingly sick wives", abandoning whatever previous categorization they have for "serious illness" in the original paper
Twenty five year marriage here. Her mental health issues started a few years in and were manageable. She did short term institutionalization later on. The end times were lots of hunkering down. She eventually left (but I still helped where I could).
Me not ending the marriage sooner - this was reasonably questioned. Some of that was me being a guy.
Having to deal with a mentally ill relative (dementia) presently, and it's no fun at all that he constantly makes stuff up and lies to everyone about everything. :(
Also no fun that he's manipulative as hell, and switches to tantrum level anger at the drop of a hat. (ugh x 10)
And everything is always someone else's fault, 100% of the time. :(
> having to deal with a mentally ill relative (dementia)
Long-term in-patient mental healthcare was a thing in the US, late 19th-early 20th century. Sanitariums were built with the best of intentions and provided the best care we knew how. Some of the care was awful because our understanding isn't what it is now. Some of the care was outstanding, even by today's standards.
The visionaries who started them died off and dollars increasingly outweighed the well-being of patients. Sanitariums eventually degenerated into hellholes and were held up as an example of [vague negatives about public mental healthcare]. In the 1980s, dollars won a total victory; many institution were shuttered. Their mentally ill patients transitioned into homelessness and here we are.
My ex needs long-term, in-patient mental healthcare. She's a vet on SS disability and has Medicaid, Medicare and VA. Even with that, long term, inpatient care isn't available for her. Wealth is the only option that could get her the help she needs.
On a tangent, the term Sanitarium is also the name of some Australian and New Zealand tax rort companies, who produce a popular breakfast cereal 'Weet-bix':
The article was published in 2015 and redacted in the same year. So for context, peer review limited the damage. However the article did enter the clickbait media.
Why might the correlation exist for heart disease but not other illnesses? Obesity comes to mind, many men become unsatisfied with their relationship when their wife sits at home getting fat. If it's not that (did they control for obesity?), I can't think of a plausible reason for this pattern of correlations so maybe the whole thing is bullshit.
The correlation might exist because it is found after the experiment was created and is merely yet another experiment hallucination (that used to be called p-hacking)
47 comments
[ 0.54 ms ] story [ 115 ms ] threadJust casually dismissing it as "fakery" removes the opportunity to learn from it.
P-hacking is so endemic and so essential to modern academia, especially in the 'softer' sciences, that the whole thing is beyond saving. It's so bad that anyone who understands statistics well enough to know that most of their peers works are bogus are much more likely to be shunned than embraced because of this knowledge which is an incredible disincentive to acquiring it.
Edit: as I can’t reply to the now dead account, the first quote explains why academia cannot keep claiming innocence on such matters. The rest of my comment was my explanation as to why nothing is being done about it. Sure, we can ask them to do better but they should already know that.
Errors will always happen; they are more likely to result in shocking, surprising results; and the press actively publicizes those results over others. The original paper got widespread press coverage, and its retraction got... Much less. You end up much dumber if you outsource your thinking to the Washington Post or Facebook.
Do you mean that the original study was fake or the second study that found the discrepancy is fake? Is the retraction fake?
The one divorce I can think of in my extended family came from the onset of heart problems. Properly speaking, it was the abortion she had that caused the divorce, not the fact that she would have died otherwise.
we could short circuit these association choices if the hosts could become unburdened while the fetus is able to live
But this is a stopgap until they can be grown as readymade adults, skipping the tedious and problematic child stage altogether.
It’s pretty clear that it will be the state who assumes receivership of the fetus and subsequent human with full constitutional rights.
It’s also a pretty easy legislative problem to say the state is transferring unwanted fetuses, despite the technology for its viability not currently existing. The outcome shifts the burden of guilt - or lack thereof - away from the parent(s), over to an immune state regardless of that state’s laws on abortion. And it shifts priorities and funding measures on creating the technology to ensure a fetus’ viability, whether that turns out to be a fool's errand or not.
This will easily bridge consensus between “camps” as the discomfort over citizens killing fetus turns into a market choice instead of a legislative debate. What will the market primarily choose to end a pregnancy
https://www.forbes.com/advisor/legal/divorce/divorce-statist...
However, my judgment was insufficient and I didn't properly recognize that sustaining the marriage was harming our kids.
In an earlier time, the harm would have been much less. Kids (my generation & earlier) once had direct, relationship-level access to a number of caring, invested adults - related and not.
My kids had the two of us 24/7 and that's about it. My wife and I had 10x the parenting demands (time) put upon us that my parents, grandparents-> adults through human history had.
This is p-hacking after-the-fact, right? Seems like the classic example: The broader hypothesis doesn’t hold, but if you look for ways to slice and dice the data you’re likely to find a (spurious?) correlation eventually.
xkcd/882
The blog post list this as the original Stata code:
> replace event`i’ = 1 if delta_mct`i’ != 0 | spouse_delta_mct`i’ != 0
and this as the correct one:
> replace event`i’ = 1 if (delta_mct`i’ != 0 | spouse_delta_mct`i’ != 0) & delta_mct`i’ != . & spouse_delta_mct`i’ != .
It looks like the authors didn't properly handle missing values in Stata, leading to marking people with missing health information being marked as being "severely ill" instead of being excluded from the analysis.
It's an unfortunate mistake, but it happens.
(Honest question)
You're saying the original analysis was wrong due to a coding error. I believe that's also true, but that's not what they were discussing. The variable names are inscrutable, but the article text also seems to imply that line (mis)codes divorce, not severe illness:
> People who left the study were actually miscoded as getting divorced.
So they actually found a correlation between severe illness and leaving the study. That's perhaps intuitive, if those people were too busy managing their illness to respond.
Although you're right that there's likely no relationship between wives developing heart conditions and subsequent divorce, there's not enough information from the article to know whether there's anything statistically meaningful about heart conditions specifically. It seems more likely that it's just statistical noise. I read that section and got the impression that the relationship is interesting but it doesn't necessarily mean anything, rather than implying that the original study still has some validity.
For another, the study doesn’t look at that many diseases, just four: cancer, heart disease, lung disease, and stroke. It’s possible the original study was doing some p-hacking in order to reach significance but these are pretty major categories of disease so they seem defensible. It would not surprise me if these are the only diseases in the original dataset.
Finally, the results are significant at the .01 level. Combined with the number of diseases, this is not anything as blatant as the classic XKCD green jellybean comic, although more subtle p-hacking could be at play.
Retracted: https://journals.sagepub.com/doi/pdf/10.1177/002214651456835...
Corrected: https://journals.sagepub.com/doi/abs/10.1177/002214651559635... (Not open access but it’s available from sci-hub.)
I take issue with this framing. Starting with saying that when women get sick, risk of divorce elevates, immediately followed by saying this happens only in one type of sickness is still very misleading. The entire first sentence is misrepresenting the actual findings.
Pure speculation, ofc ;)
Me not ending the marriage sooner - this was reasonably questioned. Some of that was me being a guy.
Ugh, that sounds familiar.
Having to deal with a mentally ill relative (dementia) presently, and it's no fun at all that he constantly makes stuff up and lies to everyone about everything. :(
Also no fun that he's manipulative as hell, and switches to tantrum level anger at the drop of a hat. (ugh x 10)
And everything is always someone else's fault, 100% of the time. :(
Hunker down indeed.
Long-term in-patient mental healthcare was a thing in the US, late 19th-early 20th century. Sanitariums were built with the best of intentions and provided the best care we knew how. Some of the care was awful because our understanding isn't what it is now. Some of the care was outstanding, even by today's standards.
The visionaries who started them died off and dollars increasingly outweighed the well-being of patients. Sanitariums eventually degenerated into hellholes and were held up as an example of [vague negatives about public mental healthcare]. In the 1980s, dollars won a total victory; many institution were shuttered. Their mentally ill patients transitioned into homelessness and here we are.
My ex needs long-term, in-patient mental healthcare. She's a vet on SS disability and has Medicaid, Medicare and VA. Even with that, long term, inpatient care isn't available for her. Wealth is the only option that could get her the help she needs.
https://en.wikipedia.org/wiki/Sanitarium_Health_and_Wellbein...
The companies are owned by a church, so even though wildly profitable don't pay any tax due to their owner being a religious organisation. :( :( :(