A local health office developed a questionnaire to assess satisfaction with social interactions, depression, self-esteem, mirror avoidance, and time spent in closets. Because of concerns about using multiple t-tests, the investigators used a Bonferroni correction; was divided by 5, so only p-values <0.01 were considered statistically significant. What effect will this have on:

a) The type 1 error rate?

b) The type II error rate?

c) Power?

d) Degrees of freedom?