> What evidence do you have that the two studies are bad?
The study indicts itself (i.e. it describes its methods, and the methods are not ones that lead to strong evidence).
The study population is only non-representative if applying it to populations that don't match. For example, this study supports the results for university students 18-22. Again, we come back to the point that just because it only applies for a subset does not make it "BS".
So what's wrong with the other study? Or will it be more of the same generalized claims about poor methods without any specific details from the study being given, nor any supporting facts to back up your determination?
When you sample a non-representative subset, you often get different results from the population at large. For example, you might get a different answer if you ask the Porsche enthusiasts club what the best car is, compared to if you ask random people in the population.
Undergrads are the same way. Asking a group of people, almost half of whom are virgins, questions about romantic relationships . . . well let's just say it might not generalize to the adult population.
This is covered in detail in any intro stats course or textbook. You may also be able to Google something like "unrepresentative subset psychology" to get an explainer. I can't do the issue complete justice in a comment here.
> So what's wrong with the other study?
I'm sorry, I did not sign up for reviewing every study you link me to. I reviewed the one, and explained my perception that it is a good example of the overall quality of the research on this topic. If you want an assessment of the other study, you'll need to look elsewhere.
It sounds a bit like trolling when you call something BS, then the source you give shows evidence that it does work in some populations. Then there are two other supportive studies. Yet you want to give non-sensical and non-applicable examples of asking Porche owners questions.