I’m very interested in the idea that most of the negative recorded effects of sleep deprivation are caused by sleepiness, not sleeplessness. However, there are some effects that you cannot attribute to sleepiness, like hormonal changes.
I would love to see more research done, but I think I’ve decided epistemically that I have to trust the body of knowledge before I trust somebody on the internet. It seems like you’re alleging an entire conspiracy in sleep science, and your evidence is not good enough to brook that.
In addition, using evidence like a sleep scientist falsifying data once to write a book does not actually help your case. Like a lot of your evidence, it’s n=1, and it adds to the perception that you’re alleging a conspiracy that really has no motive to exist.
Your example of a sleep deprivation expert one time saying that 5.5 hours can be alright is another n=1 case and it’s frustrating to read because you seem like someone who’s very passionate about this topic, but you need to write better and discriminate among evidence so that you lead with your best data.
Edit: Reversed stupidity is not brilliance. Saying that the state of sleep of sleep research is poor does not support making wild claims similarly supported by poor research.
One of Guzey’s projects is to advance his own claims about how sleep works, and I agree that his data is on the bottom of the epistemic pyramid. His other project, though, is a critique of the interpretation of sleep research data, and here he is on firmer ground.
You can check out my comment above showing that investigating a meta-study of sleep research on the type of sleep deprivation Guzey’s mainly focused on (i.e. a consistent 5.5 hours or of sleep so per night) shows that the meta-study is profoundly flawed. You can check it out for yourself—that’s how science is supposed to work!
In addition, using evidence like a sleep scientist falsifying data once to write a book does not actually help your case.
I think you should split up your assertions about Guzey’s speculations about how sleep works, his criticism of sleep studies, and his assertions about the state of the field.
Let’s say that the field really was in a poor state, with a heavy admixture of poor studies, misleading analysis, and various intensities of misrepresentation up to and including outright fraud. We’ve seen this phenomenon in other scientific fields. Our priors ought to be low, perhaps, but it’s not impossible by any means.
If we wanted to find out if this was true, how would we go about it? Starting with an epistemic spot-check of a widely praised popular book by a leading sleep researcher makes sense.
Discovering the problems Guzey found, such as trimmed graphs, uncited claims not representative of the findings of any real study, false claims about health declarations by the WHO, and an uncritical uptake by mass media, are exactly the sort of things that might naturally motivate somebody to speak up and keep researching. If this finding “doesn’t help his case,” then what could?
Guzey’s not alleging a conspiracy (“a secret plan by a group to do something unlawful or harmful”).
Edit: Actually, he did say “’m pretty sure that the entire “not sleeping ‘enough’ makes you stupid” is a 100% psyop.” It’s not clear who exactly is making this claim (sleep researchers are far more specific and formal, even if their claims are problematic in other ways), so this both has shades of a conspiracy accusation and a weakman/strawman. =
He’s alleging a history of bad behavior that results from bad incentives and selection effects:
So why has sleep research not been facing a severe replication crisis, similar to psychology?
First, compared to psychology, where you just have people fill out questionnaires, sleep research is slow, relatively expensive, and requires specialized equipment (e.g. EEG, actigraphs). So skeptical outsiders go for easier targets (like social psychology) while the insiders keep doing the same shoddy experiments because they need to keep their careers going somehow.
Second, imagine if sleep researchers had conclusively shown that sleep is not important for memory, health, etc. – would they get any funding? No. Their jobs are literally predicated on convincing the NIH and other grantmakers that sleep is important. As Patrick McKenzie notes, “If you want a problem solved make it someone’s project. If you want it managed make it someone’s job.”
you need to write better and discriminate among evidence so that you lead with your best data.
This I agree with. As I noted in another comment, Guzey’s a lot of things right here, except that the tone and format of his post makes his argument feel hard to respect. It comes across as ranty, manic, highly motivated (“passionate”), the sort of style we now associate with Qanon. But this is LessWrong, and I think that here, at least, we should try to focus on substance over style.
This I agree with. As I noted in another comment, Guzey’s a lot of things right here, except that the tone and format of his post makes his argument feel hard to respect. It comes across as ranty, manic, highly motivated (“passionate”), the sort of style we now associate with Qanon. But this is LessWrong, and I think that here, at least, we should try to focus on substance over style.
Tbh I’m very confused about the issues you have with my tone. I very deliberately called the post “Theses” on sleep! And I believe I very explicitly wrote that these are my conclusions from my reading of the literature and in no way am trying to make people think that the analogies that I thought of that I find convincing should be convincing to other people. I simply put forward some theses with extremely variable level of evidence between them but for each of these it is the case that I personally am convinced that they are actually true.
in no way am trying to make people think that the analogies that I thought of that I find convincing should be convincing to other people.
This reads as a denial of persuasive intent, which is clearly not the case.
Your post consists of more than analogies that you thought of. It also consists of data, your own self-experiments, arguments, citations, critiques of the literature and of Matthew Walker’s book. It doesn’t need to make us convinced of the truth of all your beliefs re: sleep to be a piece of writing with persuasive goals. I posit that virtually anybody who reads either this or your original piece on Walker’s book would find them clear examples of scientific writing that’s aiming to persuade the reader.
I happen to find your work pretty interesting, something I’ve already spent a lot of time investigating here in the comment section. It’s the kind of information I’d want to share and explore with other people. As you were the one who brought it to my attention, I’d like to be able to use you and your writing as a reference when I do so.
However, I have to consider what the reaction of the person I shared it with might be if I did so.
Knowing the typical reactions of people who are interested in this topic but who aren’t diehard LessWrongers, I anticipate (for reasons that are hard to make clear) that they would be turned off by stylistic features of your writing before they opened their minds enough to consider the substance of your theses in any meaningful way. Moreover, they would likely view me as foolish for reading something with those stylistic features, because there is a linguistic stereotype that those features map onto a certain type of particularly low-status person: an internet conspiracy theorist.
It’s not particularly fair or right that this stereotype exists. Consider that at one time, and to some extent still today, the accents of certain ethnic groups were/are taken as a negative indicator of intelligence. While this stereotype is harmful, it also exists, and members of these groups sometimes choose to speak in an accent deemed “higher-class” in order to elicit certain desirable reactions from groups with a tendency to stereotype. One of the terms here is “code-switching.” An example is an American Black person switching from Ebonics to white-sounding English in the classroom, and speaking Ebonics at home or with their friends.
Some of the stylistic features that I fear would be taken as sounding like a conspiracy theorist include:
Run-on sentences
Use of terms like “psyop” and Scientific Consensus (TM)
Inclusions of Twitter threads, particularly by scientific nonentities
Lots of switching between bold, italics, all-caps.
The direct address (“Even if I convinced you about the “sleeping too much” part, you are still probably wondering: but what does depression have to do with anything? Isn’t sleeping a lot good for mental health? Well…”)
Including speculations from other authors without any context for who they are, or why we’d particularly care about their speculation. Example: the Jeremy Hadfield quote.
Including irrelevant details like Nassim Taleb quote-tweeting you. If a reader didn’t know who you are, or who Taleb is, this could potentially come off as somewhere between confusing and narcissistic.
Leading a section with the assertion that “the vast majority of it being small-n, not pre-registered, p-hacked experiments,” followed not by a methodical attempt to prove the point but by further assertions about the bad motives of the researchers. [Group] is doing [bad thing] and because of [bad motives] is a basic pattern of conspiracy-theory discourse; I’m sure you can fill in the blanks yourself with examples. It’s not that no group has ever done a bad thing for bad motives, but that if you’re going to make that claim, you really need to back it up. You’ve done that with Matthew Walker, but not with his colleagues, and these are human beings we’re talking about. They deserve individual consideration on their merits and substantial evidence of wrongdoing before they’re on the receiving end of a callout, even on a blog. This is especially important because you literally accuse them of the moral equivalent of murdering tends of thousands of people.
Implicitly insulting the reader if they disagree with you (“Why is everyone in love with charlatans”)
Not all of these are individually inappropriate, but in the aggregate, they convey a strong impression that it’ll look bad to publicly take you seriously. They also make it seem like, even if you are correct on some or all of these theses about the incorrectness of current sleep research or the correctness of your own speculations, that you are operating with a disturbing attitude on how to navigate the politics. A person with a bright idea who’d be a terrible ally.
My strongest objection to your writing style is the accusation that sleep researchers in general are doing shoddy research and culpable for the equivalent of killing people. This is the point at which I would have stopped reading your article (and lowered my likelihood of reading other things you write in the future), if it weren’t for the fact that Elizabeth was the one who curated the piece.
The other points I raise here are annoying, but I could wade through them. The problem is that, as I say, they make it hard to consider sharing this piece with others, because of they way I expect the stereotype they invoke would reflect on me. If you would like to continue driving conversation on this important topic—and I think you’re onto something here and should continue—I strongly recommend that you eliminate the unsupported and hystrionic accusations, particularly the murder-like accusations and accusations against groups. I also recommend that you ask your editors to help you eliminate stylings that smack of the conspiracy theorist, since your comment makes me think that this is just not something you have a sensitivity to. That’s no sin, but it is a problem for your readability.
re: persuasive intent—yes, of course I want to persuade people but I’m believe I’m being very clear about the fact that some sections are just analogies.
My strongest objection to your writing style is the accusation that sleep researchers in general are doing shoddy research and culpable for the equivalent of killing people. This is the point at which I would have stopped reading your article (and lowered my likelihood of reading other things you write in the future), if it weren’t for the fact that Elizabeth was the one who curated the piece.
I believe this is a very strong misreading of what I wrote. I did write that sleep researchers in general are doing shoddy research (I’m pretty sure this is true). I never wrote that they are culpable for the equivalent of killing people.
Here’s the paragraph this is referring to, I believe:
Why are people not all over this? Why is everyone in love with charlatans who say that sleeping 5 hours a night will double your risk of cancer, make you pre-diabetic, and cause Alzheimer’s, despite studies showing that people who sleep 5 hours have the same, if not lower, mortality than those who sleep 8 hours? Convincing a million 20-year-olds to sleep an unnecessary hour a day is equivalent, in terms of their hours of wakefulness, to killing 62,500 of them.
I thought specifically about how to phrase the last sentence in order for it to be only about facts rather than accusing anyone and the “convincing a million 20-years-olds” appears clearly to be just a thought experiment to me.
appears clearly to be just a thought experiment to me.
The point isn’t what you intended to come across in your writing, but what actually does come across in your writing, and the expectations that creates in the reader about how others will perceive your writing.
By analogy, let’s say you go to a party and tell a joke making fun of my friend Sarah’s shoes. You think it’s funny and mean it as a bit of friendly teasing. I know that you’re a little nervous and are just trying to connect, and the joke honestly seems kind of funny to me.
However, I also know Sarah’s sensitive about her shoes, and that the others who heard the joke probably hear it as mean-spirited, because they don’t know you very well. Plus, the joke really did feel mean to me, even though I also found it humorous at the same time.
Because of that, I feel pressure to reprimand you, and maybe not to bring you back to another party in the future. This is partly because I want to make Sarah feel defended, but also because I’m concerned that others will think I’m mean if I don’t distance myself from you. They’ll certainly think that if I then go around telling other people the joke.
It’s this sort of reaction that the aspects of your writing I pointed out are provoking. And yes, in a real-life party situation, you could apologize and make amends, and things would probably be OK. However, on the internet, in a piece of writing, you don’t have that opportunity. It has to come off right the first time, without “help” from the comments, except in dialog with the very small number of highly-engaged LessWrongers here.
Because of the effort you’ve put into writing this up and seeking attention for your ideas, and also because I find your ideas intriguing, I’d prefer if you produced writing that avoided the perception problem I’m describing here. That way, I could more easily share this information and build on it over time. Right now, it’s a real barrier.
Incidentally, your math is wrong.
Life expectancy in the USA at age 20 is about 62 years.
24 hours/day * 365 days/year * 62 years (life expectancy at age 20) = 543,120 hours ~= 500,000 hours
If sleep scientists are convincing 1 million 20-year-olds to waste an extra hour on unnecessary and unwanted sleep every day, that could be interpreted as 2 “lives” per day.
It would take 85.6 years for sleep scientists to “waste” that much of this cohort of 1 million 20 year old’s time, all other assumptions granted. And I think you should probably address the fact that there are many more than 1 million people, while sleep scientists probably share far less than full responsibility for the sleep choices of the average person. Also, the difference between an extra hour of sleep and an extra hour of death is probably, to most people, vast.
Let’s say you take this criticism into account and wanted to word the sentence I found offensive. You might say something like:
Based on [calculation with appropriate caveats and qualifications], I estimate that the US adult population of 209 million people gets about 1 unnecessary hour of sleep per day. This amounts to about 24,000 person-years every day, or around 400 lifetimes. It adds up to almost 150,000 lifetimes every year, just in the USA. If that extra hour is “junk sleep,” as I contend, we are missing out on a huge amount of happiness in the pursuit of our pillows!
It’s not your style, and it’s a lot longer and less punchy. But it still conveys the scope of the issue, while framing it as a goal to be (perhaps) attained rather than an occasion for blame.
I leave it to you to figure out how to make use of this information.
This is along similar lines to criticism I sent to guzey before publication, albeit many times more eloquent. My criticism was dismissed similarly. The offhand dismissal of considered criticism also sounds like the discourse of people whose ideas I’m wary of. This in turn makes me wary of guzey’s ideas, which I am otherwise sympathetic towards.
I’m confused about your pushback to AllAmericanBreakfast’s (great) feedback on your style, which I find antagonistic to the point that (like AAB) I’m not comfortable sharing it with anyone, despite broadly agreeing with your conclusions and thinking it’s important.
> Convincing a million 20-year-olds to sleep an unnecessary hour a day is equivalent, in terms of their hours of wakefulness, to killing 62,500 of them.
I thought specifically about how to phrase the last sentence in order for it to be only about facts rather than accusing anyone and the “convincing a million 20-years-olds” appears clearly to be just a thought experiment to me.
For what it’s worth, I baulked at that sentence too. If you want to avoid the extra connotations, you could phrase it more like ”...will cause them collectively to forgo 62,500 lifetimes’ worth of waking hours”. (Hopefully something less clunky than that, though.)
edit: to clarify, my issue was with the comparison, not with the implied blame. Although you explicitly claim equivalence only ‘in terms of their hours of wakefulness’, to me as a reader it seems like you are doing one of two things with that sentence: either suggesting that causing a million 20-year olds to sleep an extra unnecessary hour per day is, in terms of overall badness, somewhere in the ballpark of killing 62,500 of them; or making the comparison for no good reason other than rhetorical shock value, knowing that it is technically defensible due to the qualifier (‘in terms of their hours of wakefulness’), but only because that allows you to brush over the differences between extra sleep and premature death.
You specify a style for citation! By your own logic, this should be of academic-level rigour, surely? Pleading ‘oh it isn’t supposed to be convincing’ is the exact same motte and bailey that Matt Walker is doing with his pop-sci that he self-cites.
This is an amazing bit of work, and one of the main reasons I come to LW is to find interesting, well-supported arguments that make me revise or at least question what I believe about important stuff. This does that, and I want to send it to everyone I know. But it’s hard to do so when you undermine your credibility at points (in basically the ways that All-American Breakfast has outlined).
People are going to be motivated to preserve their dearly held beliefs about sleep, and you give them unnecessary ammo to dismiss you as an internet crazy.
There’s a difference between a field where we have clear evidence that one of the senior scientists in the field falsified data and afterward he lost status for doing that and a field that just doesn’t care. There’s misbehavior by the people who allow Walker to give a keynote speech despite him making up a lot of claims in his book.
I’m very interested in the idea that most of the negative recorded effects of sleep deprivation are caused by sleepiness, not sleeplessness. However, there are some effects that you cannot attribute to sleepiness, like hormonal changes.
If you read the comments then guzey specially advised against changing sleep schedules by multiple hours in a single week the way that happened within that study.
Making an intervention for a single week tells you little about what happens in people with chronic exposure to the same conditions.
The p-values relevant for testosterone are on the lower side, with one them 0.049 (which screams p-hacking) and another at 0.02 (also really shitty). A reasonable back-of-the-envelope method to correct for p-hacking and publication bias involves multiplying the p-values with 20 (the reasoning is not super-involved. think about what happens to the truncated normal distribution in the case of complete publication bias); in that case, none of the testosterone-related p-values in said paper are significant. I feel comfortable ignoring it.
I’m very interested in the idea that most of the negative recorded effects of sleep deprivation are caused by sleepiness, not sleeplessness. However, there are some effects that you cannot attribute to sleepiness, like hormonal changes.
I would love to see more research done, but I think I’ve decided epistemically that I have to trust the body of knowledge before I trust somebody on the internet. It seems like you’re alleging an entire conspiracy in sleep science, and your evidence is not good enough to brook that.
In addition, using evidence like a sleep scientist falsifying data once to write a book does not actually help your case. Like a lot of your evidence, it’s n=1, and it adds to the perception that you’re alleging a conspiracy that really has no motive to exist.
Your example of a sleep deprivation expert one time saying that 5.5 hours can be alright is another n=1 case and it’s frustrating to read because you seem like someone who’s very passionate about this topic, but you need to write better and discriminate among evidence so that you lead with your best data.
Edit: Reversed stupidity is not brilliance. Saying that the state of sleep of sleep research is poor does not support making wild claims similarly supported by poor research.
One of Guzey’s projects is to advance his own claims about how sleep works, and I agree that his data is on the bottom of the epistemic pyramid. His other project, though, is a critique of the interpretation of sleep research data, and here he is on firmer ground.
You can check out my comment above showing that investigating a meta-study of sleep research on the type of sleep deprivation Guzey’s mainly focused on (i.e. a consistent 5.5 hours or of sleep so per night) shows that the meta-study is profoundly flawed. You can check it out for yourself—that’s how science is supposed to work!
I think you should split up your assertions about Guzey’s speculations about how sleep works, his criticism of sleep studies, and his assertions about the state of the field.
Let’s say that the field really was in a poor state, with a heavy admixture of poor studies, misleading analysis, and various intensities of misrepresentation up to and including outright fraud. We’ve seen this phenomenon in other scientific fields. Our priors ought to be low, perhaps, but it’s not impossible by any means.
If we wanted to find out if this was true, how would we go about it? Starting with an epistemic spot-check of a widely praised popular book by a leading sleep researcher makes sense.
Discovering the problems Guzey found, such as trimmed graphs, uncited claims not representative of the findings of any real study, false claims about health declarations by the WHO, and an uncritical uptake by mass media, are exactly the sort of things that might naturally motivate somebody to speak up and keep researching. If this finding “doesn’t help his case,” then what could?
Guzey’s not alleging a conspiracy (“a secret plan by a group to do something unlawful or harmful”).
Edit: Actually, he did say “’m pretty sure that the entire “not sleeping ‘enough’ makes you stupid” is a 100% psyop.” It’s not clear who exactly is making this claim (sleep researchers are far more specific and formal, even if their claims are problematic in other ways), so this both has shades of a conspiracy accusation and a weakman/strawman. =
He’s alleging a history of bad behavior that results from bad incentives and selection effects:
This I agree with. As I noted in another comment, Guzey’s a lot of things right here, except that the tone and format of his post makes his argument feel hard to respect. It comes across as ranty, manic, highly motivated (“passionate”), the sort of style we now associate with Qanon. But this is LessWrong, and I think that here, at least, we should try to focus on substance over style.
Tbh I’m very confused about the issues you have with my tone. I very deliberately called the post “Theses” on sleep! And I believe I very explicitly wrote that these are my conclusions from my reading of the literature and in no way am trying to make people think that the analogies that I thought of that I find convincing should be convincing to other people. I simply put forward some theses with extremely variable level of evidence between them but for each of these it is the case that I personally am convinced that they are actually true.
This reads as a denial of persuasive intent, which is clearly not the case.
Your post consists of more than analogies that you thought of. It also consists of data, your own self-experiments, arguments, citations, critiques of the literature and of Matthew Walker’s book. It doesn’t need to make us convinced of the truth of all your beliefs re: sleep to be a piece of writing with persuasive goals. I posit that virtually anybody who reads either this or your original piece on Walker’s book would find them clear examples of scientific writing that’s aiming to persuade the reader.
I happen to find your work pretty interesting, something I’ve already spent a lot of time investigating here in the comment section. It’s the kind of information I’d want to share and explore with other people. As you were the one who brought it to my attention, I’d like to be able to use you and your writing as a reference when I do so.
However, I have to consider what the reaction of the person I shared it with might be if I did so.
Knowing the typical reactions of people who are interested in this topic but who aren’t diehard LessWrongers, I anticipate (for reasons that are hard to make clear) that they would be turned off by stylistic features of your writing before they opened their minds enough to consider the substance of your theses in any meaningful way. Moreover, they would likely view me as foolish for reading something with those stylistic features, because there is a linguistic stereotype that those features map onto a certain type of particularly low-status person: an internet conspiracy theorist.
It’s not particularly fair or right that this stereotype exists. Consider that at one time, and to some extent still today, the accents of certain ethnic groups were/are taken as a negative indicator of intelligence. While this stereotype is harmful, it also exists, and members of these groups sometimes choose to speak in an accent deemed “higher-class” in order to elicit certain desirable reactions from groups with a tendency to stereotype. One of the terms here is “code-switching.” An example is an American Black person switching from Ebonics to white-sounding English in the classroom, and speaking Ebonics at home or with their friends.
Some of the stylistic features that I fear would be taken as sounding like a conspiracy theorist include:
Run-on sentences
Use of terms like “psyop” and Scientific Consensus (TM)
Inclusions of Twitter threads, particularly by scientific nonentities
Lots of switching between bold, italics, all-caps.
The direct address (“Even if I convinced you about the “sleeping too much” part, you are still probably wondering: but what does depression have to do with anything? Isn’t sleeping a lot good for mental health? Well…”)
Including speculations from other authors without any context for who they are, or why we’d particularly care about their speculation. Example: the Jeremy Hadfield quote.
Including irrelevant details like Nassim Taleb quote-tweeting you. If a reader didn’t know who you are, or who Taleb is, this could potentially come off as somewhere between confusing and narcissistic.
Leading a section with the assertion that “the vast majority of it being small-n, not pre-registered, p-hacked experiments,” followed not by a methodical attempt to prove the point but by further assertions about the bad motives of the researchers. [Group] is doing [bad thing] and because of [bad motives] is a basic pattern of conspiracy-theory discourse; I’m sure you can fill in the blanks yourself with examples. It’s not that no group has ever done a bad thing for bad motives, but that if you’re going to make that claim, you really need to back it up. You’ve done that with Matthew Walker, but not with his colleagues, and these are human beings we’re talking about. They deserve individual consideration on their merits and substantial evidence of wrongdoing before they’re on the receiving end of a callout, even on a blog. This is especially important because you literally accuse them of the moral equivalent of murdering tends of thousands of people.
Implicitly insulting the reader if they disagree with you (“Why is everyone in love with charlatans”)
Not all of these are individually inappropriate, but in the aggregate, they convey a strong impression that it’ll look bad to publicly take you seriously. They also make it seem like, even if you are correct on some or all of these theses about the incorrectness of current sleep research or the correctness of your own speculations, that you are operating with a disturbing attitude on how to navigate the politics. A person with a bright idea who’d be a terrible ally.
My strongest objection to your writing style is the accusation that sleep researchers in general are doing shoddy research and culpable for the equivalent of killing people. This is the point at which I would have stopped reading your article (and lowered my likelihood of reading other things you write in the future), if it weren’t for the fact that Elizabeth was the one who curated the piece.
The other points I raise here are annoying, but I could wade through them. The problem is that, as I say, they make it hard to consider sharing this piece with others, because of they way I expect the stereotype they invoke would reflect on me. If you would like to continue driving conversation on this important topic—and I think you’re onto something here and should continue—I strongly recommend that you eliminate the unsupported and hystrionic accusations, particularly the murder-like accusations and accusations against groups. I also recommend that you ask your editors to help you eliminate stylings that smack of the conspiracy theorist, since your comment makes me think that this is just not something you have a sensitivity to. That’s no sin, but it is a problem for your readability.
re: persuasive intent—yes, of course I want to persuade people but I’m believe I’m being very clear about the fact that some sections are just analogies.
I believe this is a very strong misreading of what I wrote. I did write that sleep researchers in general are doing shoddy research (I’m pretty sure this is true). I never wrote that they are culpable for the equivalent of killing people.
Here’s the paragraph this is referring to, I believe:
I thought specifically about how to phrase the last sentence in order for it to be only about facts rather than accusing anyone and the “convincing a million 20-years-olds” appears clearly to be just a thought experiment to me.
The point isn’t what you intended to come across in your writing, but what actually does come across in your writing, and the expectations that creates in the reader about how others will perceive your writing.
By analogy, let’s say you go to a party and tell a joke making fun of my friend Sarah’s shoes. You think it’s funny and mean it as a bit of friendly teasing. I know that you’re a little nervous and are just trying to connect, and the joke honestly seems kind of funny to me.
However, I also know Sarah’s sensitive about her shoes, and that the others who heard the joke probably hear it as mean-spirited, because they don’t know you very well. Plus, the joke really did feel mean to me, even though I also found it humorous at the same time.
Because of that, I feel pressure to reprimand you, and maybe not to bring you back to another party in the future. This is partly because I want to make Sarah feel defended, but also because I’m concerned that others will think I’m mean if I don’t distance myself from you. They’ll certainly think that if I then go around telling other people the joke.
It’s this sort of reaction that the aspects of your writing I pointed out are provoking. And yes, in a real-life party situation, you could apologize and make amends, and things would probably be OK. However, on the internet, in a piece of writing, you don’t have that opportunity. It has to come off right the first time, without “help” from the comments, except in dialog with the very small number of highly-engaged LessWrongers here.
Because of the effort you’ve put into writing this up and seeking attention for your ideas, and also because I find your ideas intriguing, I’d prefer if you produced writing that avoided the perception problem I’m describing here. That way, I could more easily share this information and build on it over time. Right now, it’s a real barrier.
Incidentally, your math is wrong.
Life expectancy in the USA at age 20 is about 62 years.
24 hours/day * 365 days/year * 62 years (life expectancy at age 20) = 543,120 hours ~= 500,000 hours
If sleep scientists are convincing 1 million 20-year-olds to waste an extra hour on unnecessary and unwanted sleep every day, that could be interpreted as 2 “lives” per day.
62,500 lives / (365 days/year * 2 lives/day) = 62,500 lives / (730 lives/year) = 85.6 years.
It would take 85.6 years for sleep scientists to “waste” that much of this cohort of 1 million 20 year old’s time, all other assumptions granted. And I think you should probably address the fact that there are many more than 1 million people, while sleep scientists probably share far less than full responsibility for the sleep choices of the average person. Also, the difference between an extra hour of sleep and an extra hour of death is probably, to most people, vast.
Let’s say you take this criticism into account and wanted to word the sentence I found offensive. You might say something like:
It’s not your style, and it’s a lot longer and less punchy. But it still conveys the scope of the issue, while framing it as a goal to be (perhaps) attained rather than an occasion for blame.
I leave it to you to figure out how to make use of this information.
This is along similar lines to criticism I sent to guzey before publication, albeit many times more eloquent. My criticism was dismissed similarly. The offhand dismissal of considered criticism also sounds like the discourse of people whose ideas I’m wary of. This in turn makes me wary of guzey’s ideas, which I am otherwise sympathetic towards.
I’m confused about your pushback to AllAmericanBreakfast’s (great) feedback on your style, which I find antagonistic to the point that (like AAB) I’m not comfortable sharing it with anyone, despite broadly agreeing with your conclusions and thinking it’s important.
For what it’s worth, I baulked at that sentence too. If you want to avoid the extra connotations, you could phrase it more like ”...will cause them collectively to forgo 62,500 lifetimes’ worth of waking hours”. (Hopefully something less clunky than that, though.)
edit: to clarify, my issue was with the comparison, not with the implied blame. Although you explicitly claim equivalence only ‘in terms of their hours of wakefulness’, to me as a reader it seems like you are doing one of two things with that sentence: either suggesting that causing a million 20-year olds to sleep an extra unnecessary hour per day is, in terms of overall badness, somewhere in the ballpark of killing 62,500 of them; or making the comparison for no good reason other than rhetorical shock value, knowing that it is technically defensible due to the qualifier (‘in terms of their hours of wakefulness’), but only because that allows you to brush over the differences between extra sleep and premature death.
You specify a style for citation! By your own logic, this should be of academic-level rigour, surely? Pleading ‘oh it isn’t supposed to be convincing’ is the exact same motte and bailey that Matt Walker is doing with his pop-sci that he self-cites.
This is an amazing bit of work, and one of the main reasons I come to LW is to find interesting, well-supported arguments that make me revise or at least question what I believe about important stuff. This does that, and I want to send it to everyone I know. But it’s hard to do so when you undermine your credibility at points (in basically the ways that All-American Breakfast has outlined).
People are going to be motivated to preserve their dearly held beliefs about sleep, and you give them unnecessary ammo to dismiss you as an internet crazy.
Psyop is a word that you find in Qanon discourse but seldom in mainstream discourse.
There’s a difference between a field where we have clear evidence that one of the senior scientists in the field falsified data and afterward he lost status for doing that and a field that just doesn’t care. There’s misbehavior by the people who allow Walker to give a keynote speech despite him making up a lot of claims in his book.
If you read the comments then guzey specially advised against changing sleep schedules by multiple hours in a single week the way that happened within that study.
Making an intervention for a single week tells you little about what happens in people with chronic exposure to the same conditions.
About that paper.
The p-values relevant for testosterone are on the lower side, with one them 0.049 (which screams p-hacking) and another at 0.02 (also really shitty). A reasonable back-of-the-envelope method to correct for p-hacking and publication bias involves multiplying the p-values with 20 (the reasoning is not super-involved. think about what happens to the truncated normal distribution in the case of complete publication bias); in that case, none of the testosterone-related p-values in said paper are significant. I feel comfortable ignoring it.