Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
Homaniac
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
Homaniac
11y ago
Yes, but over sampling (over powering) a study causes issues as well due to increasing the chance of Type I Error (i.e. detecting false positives).
2.
▲
by
Homaniac
11y ago
Ah yeah, it definitely cannot be generalized to US marriages at large due to the sample only being from a limited area. Confounding factors may definitely be an issue as well.
3.
▲
by
Homaniac
11y ago
Why are you skeptical due to the sample size of 94? The sample size only needs to be large enough to capture the variance in the response for the population. The sample size does not need to increase as the population being studied increas