> We must also always consider that there could be “third variables” that cause both of the first two variables to rise. In this case, one plausible confounding

> We must also always consider that there could be “third variables” that cause both of the first two variables to rise. In this case, one plausible confounding third variable is permissive parenting. Perhaps permissive parents (in each country) simultaneously do two things: they give their kids smartphones at very young ages, and they also give them few boundaries and little structure, which then interferes with development and produces struggling young adults. From the original article. I wonder if this variable is controlled, how much of the correlation will be gone.