B: the criminal tendencies cause people to watch more TV as children

    Incorrect.

This one is tempting, but you must resist!  The question above notes an relationship between childhood television watching behavior and adult crime rate (of the former children).  This relationship is a positive one, meaning that if a child watched a lot of TV as a child, he or she would be more likely to commit a crime than a person who watched less TV as a child.  This is a correlational method, because an association was measured.  When we have a correlation, we only know there is a relationship between variables; we do not know which variable is causing the relationship.  This problem is called reverse causation.  In this case, choice B states that criminal tendencies cause people to watch more TV as children.  Isn't it possible that TV caused learning of criminal behavior too?  We don't know, so we cannot state it either way.  Causation can not be proven! Rather, we just state there is an association or relationship.