A: that watching TV as a child causes criminal behavior

    Incorrect.

In this question, a correlational method is suggested.  The study described in the question appears to have observed individuals at childhood, and compared them to themselves as adults.  Since variables were not manipulated, there was no experimental control, and there was no random assignment, this was NOT an experimental method.  Since this is a correlational rather than an experimental, causation can NOT be inferred.  When you see the words "cause" or "prove" you should recognize this answer is not correct.