>I don't know. It's a danish study, there could be regional differences or it just speaks to the type of people willing to be study subjects.
25% is too significant to hand wave away. It's a major fundamental flaw in the paper and calls into question their entire methodology
>>diet and socio economic status affects hormone levels? To what extent? Yes, not to a that great an extent.
Quantify. We're dealing with a quantified drop here, we need to use real numbers to look at the confounding effects. Provide a source with quantification. Is it 2%, 10%, 50%? This is important.
>>Can this be quantified? I don't think so.
If you can't quantify it then on what basis do you ignore it? You have no basis to ignore it, and the research is compromised.
>Where are you going with these passive aggressive comments?
Don't take it personally, I criticize research for a living, I'm just doing what I always do. This research has significant flaws I'm just asking about them. The fact that you're getting emotional in response to questions about data indicates you're not impartial.
> Do you disagree with the trend that the users with the most use had the lowest levels of testosterone and the smallest size? Is that not a fair association, on top of the differences in mean?
I have several concerns with this paper, and it does not convince me on its own, no. Is that what this picture shows you? Testis size is a secondary measure, an indirect way to assess gonadal status. What you really are concerned with is serum free test level. Why didn't you use this picture initially? I'm curious.
Attached: journal.pone.0161208.g002.png (3837x2856, 203K)