The hormone levels are unlikely to be affected by cultural biases but that's not the potential weakness I'm talking about in these studies. It's during the observation, through the eyes of the parents and scientists, of what toys the children are playing with where expectations can affect the results. On top of that, you've got biases affecting what studies get published and which of those then make headlines and get repeated in blog posts like this one.