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

Too many variables to isolate on this study to provide proof that correlation equals causation. Even with the isolating they've tried to do with the 95% CI the range is:

1 Person Office: 1 (They didn't report a range here)

2 Person Office: 1.13-1.98 sick days

3-6 Person Office: 1.08-1.73 sick days (LOWER than 2)

6+ Person Office: 1.30-2.02 sick days

Great item to study further. This isn't scientific proof of anything. As a solo worker you may come in to the office when sick as well since you don't have to worry about getting anyone else sick and so many other variables they didn't control for make this interesting but far from proof.

Direct PDF link to full study: http://www.sjweh.fi/download.php?abstract_id=3167&file_nro=1




There is a perfect positive correlation between smoking + alcohol consumption and health!

1 person: heavy smokers 15%, drinkers 16%

3-6 pers: heavy smokers 13%, drinkers 16%, sick leave +36%

2 pers: heavy smokers 10%, drinkers 15%, sick leave +50%

> 6 pers: heavy smokers 9%, drinkers 13%, sick leave +62%

Please drink and smoke more to fight sickness.


They said they controlled for those confounding variables. I bet the raw data used in this study also says smokers and drinkers are more often sick than the others.


Note that they don’t report a confidence interval for the one-person office because it is the baseline for the other rate ratios. (This is generally how the baseline group is identified when reporting ratios in journal articles.)


Assuming social conflict as well as physical health is a driver, the local peak at 2 makes perfect sense.


You've quoted rates not absolute values.




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: