7 ms·

I think the research behind this news is quite weak: only 9 days and not that many participants (18 men and 18 women).

by kuu 8y ago

xster 8y ago
This needs to be top comment. Society's tolerance for bombastic 'journalistic' claims with no consequences is the source of so much of our present issues. Conclusions are cheap.
- omeid2 8y ago
  > ...source of so much of our present issues. and past issues.
- FooHentai 8y ago
  >is the source of so much of our present issues. Citation needed. Because hyperbole is no different in meta.
  - tonyedgecombe 8y ago
    >Because hyperbole is no different in meta Citation need... Oh wait, this is discussion, not an academic paper.
    - waterhouse 8y ago
      I've seen people say this a few times, and I think it is attacking a misreading of the comment (snarkily, to boot). Charitable interpretation: In a discussion, if someone says "Citation needed", it means "I'm skeptical and I'm asking for a citation" (which is quite reasonable), not "You should have supplied a citation before I asked for it" (which I would agree is unreasonable for statements one doesn't expect to be contentious).
    - presscast 8y ago
      > Oh wait, this is discussion, not an academic paper. It's literally an academic discussion insofar as it's about the technicalities of an academic paper. So yes, you really do need to support your arguments if you want them to be taken seriously.
  - cobbzilla 8y ago
    The replication problem has been well documented, at least for the social sciences. A cursory search for “replication crisis” will yield many results. For example https://en.wikipedia.org/wiki/Replication_crisis https://en.wikipedia.org/wiki/Replication_crisis
    - Retra 8y ago
      The replication crisis is a problem, but it is almost certainly not the main source of societies ills.
  - 781 8y ago
    Here is your citation: https://www.sciencealert.com/replication-results-reproducibility-crisis-science-nature-journals https://www.sciencealert.com/replication-results-reproducibi...
wilg 8y ago
"9 days" really undersells that this is a study about the effect of weekends on sleep and they only tested one weekend.
- yoz-y 8y ago
  Yeah, I haven’t read the article and after seeing this there is really no point. If one data point is what it gets to publish a study...
cnaa 8y ago
A thesis that journalism is the continued education for adults was posted recently. I won't name the originator of that idea. Education here: Your 60 work hours during the week are fine, you don't need more sleep. Doesn't Bezos own the WP?
YayamiOmate 8y ago
I wanted to ask whether this is a serious claim or one of those "might alter the risk" articles. Seems it even worse. Thanks for sparing many people's time with your comment.
presscast 8y ago
>18 men and 18 women This very much depends on statistical power. If the effect size is big enough, 36 participants can be plenty. I suspect the effect sizes of sleep deprivation are probably quite large.
sz4kerto 8y ago
Please support your argument with numbers. 36 participants can be plenty or can be absolutely insignificant. My partner is just in the process of getting subjects for an experiment and she's trying to estimate the required number of participants. In their case 30 participants would make the experiment useless, 40-42 would make it very powerful (significant). You can't judge the weakness of an experiment just based on the number of participants. Your comment might mislead others with not enough knowledge of stats I think.
- c1as 8y ago
  Academia has had their chance. No one believes "studies" about personal life choices any more if the results conflict with each other every 5 years. It's just publication pressure (or worse, as other commenters have hinted at).
  - sz4kerto 8y ago
    > Academia has had their chance. You can't say that. (Ok, of course everyone has the right to her opinions.) The point of the scientific method is that it is, in many ways, self-correcting. It's getting addressed, albeit slowly. It will be counterproductive for all of us if we simply dismiss all results we're not comfortable with that it's only publication pressure. I mean -- you can say that personally, and it probably makes sense, but as a society we need to get social sciences and medicine into a better shape instead of just 'letting it go'.
    - pelasaco 8y ago
      > It will be counterproductive for all of us if we simply dismiss all results we're not comfortable with that it's only publication pressure. That's not what he said. Just those which "each 5 years" have completely different results.
  - Ragnarork 8y ago
    > Academia has had their chance Oh, so what should we go to now in order to try to expand our knowledge about this? What are you suggesting?
rat9988 8y ago
I don't think you can dismiss a scientific article without arguments. Please prove your points.
- dkersten 8y ago
  A 9 day study includes only one weekend. In a study about weekends. One data point is never enough to make any inferences.
  - rat9988 8y ago
    It depends what you are trying to prove. To prove that catchup sleep doesn't work, you need to find at which weekend it stops working. If it's the first week end, you don't need to iterate, as the person's condition will only get worse.
    - dkersten 8y ago
      > If it's the first week end, you don't need to iterate, as the person's condition will only get worse. Then you let your subjects rest up and repeat it again another few times, to make sure that its repeatable and not a coincidence.
el_cujo 8y ago
I also take issue with how the article mentions that sleep is not a balance sheet, but doesn't really lay out that this was explicitly tested. They said the participants slept only five hours a night for weekdays. 8 hours is the generally recommended amount for a good nights sleep, so to truly bring balance to that schedule, they'd have a 15 hour sleep-debt to make up on the weekends, in addition to the full 8 hours they'd need to sleep those two weekend days on their own. I seriously doubt the participants were able to sleep 16 hours on both friday and saturday night to fully pay the sleep debt even at their level of sleep deprivation, so I don't think you can really say the "balance sheet" approach is wrong. I'm not necessarily saying I support a balance sheet view of sleep, I just think these kinds of articles are not very careful with the kind of wording they use. Also maybe I'm just living under a rock, but I don't feel like its that common to only get 5 hours of sleep every single weekday unless you're like a single parent with a baby (in which case you probably wouldn't magically get 8 hours on the weekend), I'd be more interested to see what happens if this was done in smaller spurts, like two consecutive days with three or four hours of sleep to emulate some work or school deadline approaching, but then 8 or 9 the rest of the week after its complete. As it stands now, the title of the article comes off as click bait because its a pretty fear-mongering conclusion drawn from what I consider to be pretty extreme circumstances.
- wott 8y ago
  > They said the participants slept only five hours a night for weekdays. 8 hours is the generally recommended amount for a good nights sleep, so to truly bring balance to that schedule, they'd have a 15 hour sleep-debt to make up on the weekends, in addition to the full 8 hours they'd need to sleep those two weekend days on their own. I seriously doubt the participants were able to sleep 16 hours on both friday and saturday night to fully pay the sleep debt even at their level of sleep deprivation, so I don't think you can really say the "balance sheet" approach is wrong. Well, if you take a reference of 7 hours instead of 8, it is more possible. That's kind of what I was doing when I was in high-school: sleeping 5 hours or under on week days, so a total of about 24-25 hours over 5 days; and then sleeping 12 hours or more on week-end nights. So I was sleeping as much in those 2 days, as I was in the other 5, for a total of about 50 hours / week, which is around 7 hours per day on average. Note that I do not recommend this rhythm! Periodically during those years, I suffered from awful 3- to 5-days long periods during which I was... I don't know how to explain, like if there was a cotton wall between the world and me, I had no control over anything, everything was happening automatically, I had kind of visions of stuff that would happen not matter what I'd do, etc. I guess all of this was caused by sleep deprivation. Gee, I don't miss those periods... luckily it stopped after high-school. But all my life, I have had huge difficulties to fall asleep at night, and also suffered from stress, which doesn't combine well with the necessities of school times, work times, and other mandatory times: I generally managed to sleep earlier when there was no need to wake up early... If I ever need to wake up at 5 or 4 AM, I won't even manage to get any sleep (or 45 minutes with 10 sweaty 180 heartbeats per minute wake-up inside).

I think the research behind this news is quite weak: only 9 days and not that many participants (18 men and 18 women).

by kuu 8y ago

I think the research behind this news is quite weak: only 9 days and not that many participants (18 men and 18 women).

This needs to be top comment. Society's tolerance for bombastic 'journalistic' claims with no consequences is the source of so much of our present issues. Conclusions are cheap.

omeid2 8y ago
> ...source of so much of our present issues. and past issues.
FooHentai 8y ago
>is the source of so much of our present issues. Citation needed. Because hyperbole is no different in meta.
- tonyedgecombe 8y ago
  >Because hyperbole is no different in meta Citation need... Oh wait, this is discussion, not an academic paper.
  - waterhouse 8y ago
    I've seen people say this a few times, and I think it is attacking a misreading of the comment (snarkily, to boot). Charitable interpretation: In a discussion, if someone says "Citation needed", it means "I'm skeptical and I'm asking for a citation" (which is quite reasonable), not "You should have supplied a citation before I asked for it" (which I would agree is unreasonable for statements one doesn't expect to be contentious).
  - presscast 8y ago
    > Oh wait, this is discussion, not an academic paper. It's literally an academic discussion insofar as it's about the technicalities of an academic paper. So yes, you really do need to support your arguments if you want them to be taken seriously.
- cobbzilla 8y ago
  The replication problem has been well documented, at least for the social sciences. A cursory search for “replication crisis” will yield many results. For example https://en.wikipedia.org/wiki/Replication_crisis https://en.wikipedia.org/wiki/Replication_crisis
  - Retra 8y ago
    The replication crisis is a problem, but it is almost certainly not the main source of societies ills.
- 781 8y ago
  Here is your citation: https://www.sciencealert.com/replication-results-reproducibility-crisis-science-nature-journals https://www.sciencealert.com/replication-results-reproducibi...

wilg 8y ago

"9 days" really undersells that this is a study about the effect of weekends on sleep and they only tested one weekend.

yoz-y 8y ago
Yeah, I haven’t read the article and after seeing this there is really no point. If one data point is what it gets to publish a study...

cnaa 8y ago

A thesis that journalism is the continued education for adults was posted recently. I won't name the originator of that idea. Education here: Your 60 work hours during the week are fine, you don't need more sleep. Doesn't Bezos own the WP?

YayamiOmate 8y ago

I wanted to ask whether this is a serious claim or one of those "might alter the risk" articles. Seems it even worse. Thanks for sparing many people's time with your comment.

presscast 8y ago

>18 men and 18 women This very much depends on statistical power. If the effect size is big enough, 36 participants can be plenty. I suspect the effect sizes of sleep deprivation are probably quite large.

sz4kerto 8y ago

Please support your argument with numbers. 36 participants can be plenty or can be absolutely insignificant. My partner is just in the process of getting subjects for an experiment and she's trying to estimate the required number of participants. In their case 30 participants would make the experiment useless, 40-42 would make it very powerful (significant). You can't judge the weakness of an experiment just based on the number of participants. Your comment might mislead others with not enough knowledge of stats I think.

c1as 8y ago
Academia has had their chance. No one believes "studies" about personal life choices any more if the results conflict with each other every 5 years. It's just publication pressure (or worse, as other commenters have hinted at).
- sz4kerto 8y ago
  > Academia has had their chance. You can't say that. (Ok, of course everyone has the right to her opinions.) The point of the scientific method is that it is, in many ways, self-correcting. It's getting addressed, albeit slowly. It will be counterproductive for all of us if we simply dismiss all results we're not comfortable with that it's only publication pressure. I mean -- you can say that personally, and it probably makes sense, but as a society we need to get social sciences and medicine into a better shape instead of just 'letting it go'.
  - pelasaco 8y ago
    > It will be counterproductive for all of us if we simply dismiss all results we're not comfortable with that it's only publication pressure. That's not what he said. Just those which "each 5 years" have completely different results.
- Ragnarork 8y ago
  > Academia has had their chance Oh, so what should we go to now in order to try to expand our knowledge about this? What are you suggesting?

rat9988 8y ago

I don't think you can dismiss a scientific article without arguments. Please prove your points.

dkersten 8y ago
A 9 day study includes only one weekend. In a study about weekends. One data point is never enough to make any inferences.
- rat9988 8y ago
  It depends what you are trying to prove. To prove that catchup sleep doesn't work, you need to find at which weekend it stops working. If it's the first week end, you don't need to iterate, as the person's condition will only get worse.
  - dkersten 8y ago
    > If it's the first week end, you don't need to iterate, as the person's condition will only get worse. Then you let your subjects rest up and repeat it again another few times, to make sure that its repeatable and not a coincidence.

el_cujo 8y ago

I also take issue with how the article mentions that sleep is not a balance sheet, but doesn't really lay out that this was explicitly tested. They said the participants slept only five hours a night for weekdays. 8 hours is the generally recommended amount for a good nights sleep, so to truly bring balance to that schedule, they'd have a 15 hour sleep-debt to make up on the weekends, in addition to the full 8 hours they'd need to sleep those two weekend days on their own. I seriously doubt the participants were able to sleep 16 hours on both friday and saturday night to fully pay the sleep debt even at their level of sleep deprivation, so I don't think you can really say the "balance sheet" approach is wrong. I'm not necessarily saying I support a balance sheet view of sleep, I just think these kinds of articles are not very careful with the kind of wording they use. Also maybe I'm just living under a rock, but I don't feel like its that common to only get 5 hours of sleep every single weekday unless you're like a single parent with a baby (in which case you probably wouldn't magically get 8 hours on the weekend), I'd be more interested to see what happens if this was done in smaller spurts, like two consecutive days with three or four hours of sleep to emulate some work or school deadline approaching, but then 8 or 9 the rest of the week after its complete. As it stands now, the title of the article comes off as click bait because its a pretty fear-mongering conclusion drawn from what I consider to be pretty extreme circumstances.

wott 8y ago
> They said the participants slept only five hours a night for weekdays. 8 hours is the generally recommended amount for a good nights sleep, so to truly bring balance to that schedule, they'd have a 15 hour sleep-debt to make up on the weekends, in addition to the full 8 hours they'd need to sleep those two weekend days on their own. I seriously doubt the participants were able to sleep 16 hours on both friday and saturday night to fully pay the sleep debt even at their level of sleep deprivation, so I don't think you can really say the "balance sheet" approach is wrong. Well, if you take a reference of 7 hours instead of 8, it is more possible. That's kind of what I was doing when I was in high-school: sleeping 5 hours or under on week days, so a total of about 24-25 hours over 5 days; and then sleeping 12 hours or more on week-end nights. So I was sleeping as much in those 2 days, as I was in the other 5, for a total of about 50 hours / week, which is around 7 hours per day on average. Note that I do not recommend this rhythm! Periodically during those years, I suffered from awful 3- to 5-days long periods during which I was... I don't know how to explain, like if there was a cotton wall between the world and me, I had no control over anything, everything was happening automatically, I had kind of visions of stuff that would happen not matter what I'd do, etc. I guess all of this was caused by sleep deprivation. Gee, I don't miss those periods... luckily it stopped after high-school. But all my life, I have had huge difficulties to fall asleep at night, and also suffered from stress, which doesn't combine well with the necessities of school times, work times, and other mandatory times: I generally managed to sleep earlier when there was no need to wake up early... If I ever need to wake up at 5 or 4 AM, I won't even manage to get any sleep (or 45 minutes with 10 sweaty 180 heartbeats per minute wake-up inside).