Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I heard that these studies are, like all social studies, to be taken with a grain of salt. They don't replicate well.


I don't know about that. It should be fairly easy to study (and hence replicate).

1. Find a group of animals (preferably humans, but dogs, rats, pigeons, etc. should also work just fine).

2. Target a specific unnecessary behavior pattern (hobby) that all individual engage in.

3. Measure the baseline rate of that activity.

4. Introduce a reward:

4a. For a randomly selected subset the reward follows imminently the targeted behavior.

4b. For the rest they'll receive the "reward" equally often but at a time independent of the targeted behavior.

5. Drop the reward.

6. Let some time pass.

7. Measure the drop/increase in the behavior from the baseline measure.

---

[edit] format.


[flagged]


I think he's referring to the replication crisis:

https://en.wikipedia.org/wiki/Replication_crisis


Sorry, my advice is, Google for the whole replication crisis thing. I think I read something about this specific set of studies on slate star codex, so that's a good place to look.

Please pardon that I don't have enough time to look for the sources myself.


No, he's right. I wouldn't have even mentioned Social sciences. Medical studies are just as terrible. We need to question every study. Repeat said studies. Make sure the repeat studies were done right. Then study it again by outsiders.




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

Search: