Hacker News new | past | comments | ask | show | jobs | submit login

I'm bad at reading scientific papers, could someone clarify this please? It says in the results:

> Compared with nonconsumers, consumers of various amounts of unsweetened coffee (>0 to 1.5, >1.5 to 2.5, >2.5 to 3.5, >3.5 to 4.5, and >4.5 drinks/d) had lower risks for all-cause mortality after adjustment for lifestyle, sociodemographic, and clinical factors, with respective hazard ratios of 0.79 (95% CI, 0.70 to 0.90), 0.84 (CI, 0.74 to 0.95), 0.71 (CI, 0.62 to 0.82), 0.71 (CI, 0.60 to 0.84), and 0.77 (CI, 0.65 to 0.91); the respective estimates for consumption of sugar-sweetened coffee were 0.91 (CI, 0.78 to 1.07), 0.69 (CI, 0.57 to 0.84), 0.72 (CI, 0.57 to 0.91), 0.79 (CI, 0.60 to 1.06), and 1.05 (CI, 0.82 to 1.36).

It explicitly says that the all-cause mortality for unsweetened coffee was lower, that's pretty clear. But then it just gives (Confidence Interval?) numbers for "the respective estimates for consumption of sugar-sweetened coffee" but doesn't say if it lowers all-cause mortality, is that implied as well? Or are those numbers just confidence intervals without giving an assessment?




You're not bad at reading science papers, I suspect you have not picked up the techniques to read science papers. Imagine you're staring at a long and complex Bash script. And you don't know Bash but you do understand other programming languages. You understand the shape of what's there, but not the details.

Now also imagine that the self-same Bash script is written in the tersest style possible without any consideration to how accessible it might be for the reader. That's most science papers.

Once you understand how a paper is to be read, and different papers, e.g. medical, or observational (like this study) or computer science, will have differing common styles between them.

So you're not bad at reading science papers. It is simply that nobody bothered to show you the techniques yet.


tersest summary: jargon.


The hazard ratio being <1 means that mortality was lower.


I see, thanks! So those are the hazard ratios for each of the segments of consumption of coffee, right? (5 coffee drinking segments, 5 hazard ratios)


Correct, ">3.5 to 4.5" / "0.71 (CI, 0.60 to 0.84)" means "we're 95% confident that people who had 3.5 to 4.5 unsweetened coffees per day had 60-84% the risk of mortality of baseline (average 71%)".


Ohhhhh, thank you! That makes sense. It seems there was an implicit assumption of 95% (p<0.05), is that right?


Right. 95% is the most common threshold for a CI.


Lower than general population or lower than the control group?


There's also a wide range of sugar amounts you can add to coffee. I add only about 1/3 cube of sugar to a cup of coffee to slightly offset the bitterness, where I see other people sometimes toss in 3 cubes (which tastes awful to me). That's about an order of magnitude difference right there.


> I'm bad at reading scientific papers

Reading scientific papers is hard. They are dense and packed with numbers. Nobody is "good" at it, it takes effort to unpack them and digest them.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: