It is suspected that watching violence on television may have an adverse impact on human emotional development and social interaction. As part of a study on the issue, a number of young children are shown a nonviolent cartoon on one occa- sion and a cartoon containing violence on another occasion. After each cartoon viewing, each child was introduced to a group of children of similar age (who watched neither cartoon) and their interaction with the group was studied. The number of times in which it was judged that a child socialized more constructively after the nonviolent cartoon than after the violent cartoon was assessed by an assessor who had no knowledge of the type of cartoon that had been shown.

a. What is the null hypothesis in this experiment? Is the null hypothesis appropriate? What is the alternative hypothesis in this experiment? Is a one-tail or two-tail test appropriate?

b. How plausible do you personally feel each of these hypotheses are?

c. In making a decision between these hypotheses, what do you see as the costs in making a Type I error? What do you see as the costs in making a Type II error?

d. What p value do you feel is appropriate to just make you change your mind between the two hypotheses?

e. Say that the experiment was actually performed on twelve children with the following results: Five children socialized equally well after both cartoons, six children socialized better after the nonviolent cartoon than after the violent cartoon, and one child socialized better after the violent cartoon than after the nonviolent cartoon. Perform an appropriate statistical test and find the p value. Do you now believe in H0 or Ha? POLI967