The problem isn’t having variables you can’t control. The problem is not recognizing these variables, or recognizing them and being too lazy to control for them

The problem isn’t having variables you can’t control. The problem is not recognizing these variables, or recognizing them and being too lazy to control for them, and then drawing unwarranted, overly broad hypotheses. There’s a reason psych has a bigger reproducibility problem than any other field.