[Homaniac]

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 increases. Look into the concept of statistical power.

[avsbst]

I think the original commentor's skepticism, while possibly too concerned with sample size, is still valid. This study is trying to draw connections between two large and hard to understand systems, marriage and cardiovascular health. While the sample size may have captured enough variance in a certain population to generalize the results I doubt they can be generalized world wide, nation wide, state wide or possibly even city wide.

For example, how did they find these participants? Did the researchers just pull married couples from the BYU campus or from across the city? Given BYU and Utah's population are they mostly LDS couples? What about their race given that the LDS is significantly more caucasian than the US overall? What about diet given that LDS members generally don't consume alcohol or caffeinated beverages? What about geography? Perhaps temperature or altitude affected the cardiovascular markers or measurements were taken under different conditions for different groups?

All of the above things I've mentioned may or may not affect cardiovascular health, and I doubt they captured or had the time and resources to control for all these factors and any additional ones I didn't bring up. So I think being skeptical is okay. Until this study has been reproduced multiple times or we understand the entire pathway connecting marriage and heart health (a highly unlikely occurrence) I will remain skeptical because that's how good science works.

[Homaniac]

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.

[lutorm]

This is only true if you know that you are sampling from the underlying population fairly, and you can't conclude that from just looking at the sample variance unless you know the distribution you are sampling from.

[tertius]

> The sample size only needs to be large enough to capture the variance in the response for the population.

And you need a representative sample to capture the variance of a population. The smaller the sample, the less likely you will be in capturing the actual variance.

[Homaniac]

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).

[tertius]

I agree and that's not really what I'm saying.

I think if you want to generalize (which is very rarely done in science) then the sample size has to be representative of your (general) population.

Now we know that media (i.e. NYT) likes to take the extreme specificity that makes studies significant and apply it broadly. That's really where my contention comes from.