Hacker News Reader

Comment by Markoff | Hacker News Reader

Meanigless study with only 21 babies, extremely small sample size to make some general statements.

> Results showed that the most effective way to put a baby to sleep was to carry them and walk around for five minutes, sit and wait for five to eight minutes, then put them to bed.

Also this is nice theory, we all know of course if you hold baby (and eventually walk) the baby can fall asleep, the problem is to put the baby down without waking it up, I would like to see this quantified over hundreds of children how many times will baby wake up after following these steps and then putting them in bed.

View on HN

1. It's an exploratory study which both confirms previous results and builds on them. Furthermore, a sample size of 21 is not too inherently small to draw conclusions - that would depend on the effect sizes between conditions...

2. The study is explicitly about whether or not the children stay asleep after the different proposed routines. This is one of the main points of the article and is directly addressed within:

"When an infant falls asleep in one’s arms or in a stroller, parents would then want to put the infant to bed. However, this laydown procedure often makes the infant alert again. In this study, 9 of 26 (34.6%) sleeping infants awoke by 20 s after the laydown, while the remaining 17 infants stayed asleep. "

https://www.cell.com/current-biology/pdf/S0960-9822(22)01363...

Why comment with a critique if you didn't even have a cursory look at the paper, or even understand what the dependant variables are?

> Why comment with a critique if you didn't even have a cursory look at the paper, or even understand what the dependant variables are?

Because I am commenting on article posted on HN. No such information provided in linked article, nor is the study linked there.

> In this study, 9 of 26 (34.6%) sleeping infants awoke by 20 s after the laydown, while the remaining 17 infants stayed asleep.

Thanks for link to the study and thanks for proving me right with the quote. More than 1/3 of children waking up almost instantly within 20 seconds just confirms this is hardly usable and "best" way to put crying baby to sleep even with such small sample size. And if we had more data I'd expect most of the babies woke up shortly after that 20 seconds limit. I tried to look at the charts but they are difficult to read, since they don't provide normal table with times for all babies.

It's similar nonsense as road casualties, which are not comparable across countries since some countries count causalty as someone dying after car accident within 7 or 30 days and if you crash and lay in coma 31 days you are suddenly not road casualty anymore, same as here amusing 20 seconds threshold for success.

I'm getting quite depressed by the increase prevelence of these types of dismissive comments on studies. You see this exact argument posted especially on Reddit.

The second increasingly common popular comment is "Yeah but is it causal? and not correlation?", including popular follow ups such as did they correct for socio economic status? (Which in the vast majority of studies, they always do)

I'm curious to know if the posters of these comments have been to univsersity / college? Have they taken any classes on the basics of scientific method? I remember my bachelors had a course on it, at least.

It's great that more people are interested in science, but it's not great that people dismiss entire studies because they do not understand how science works.

atleast read the link before commenting nonsense.

Reply on news.ycombinator.com