1. What is a Type I error?
    We reject H0 when it is actually true. it equals to alpha
  2. What is a Type II error
    we accept H0 when the alternative is true
  3. hypothesis testing: what is statistical power of a test?
    power is probability of rejecting null hypothesis when null hypotehsis is false. power equal to 1 - beta.
  4. hypothesis testing: general procedure for obtaining a decision rule?
    assume H0 is true, probability distribution of test statistic, set alpha.
  5. decreasing level of significance for test of hypothesis change the probability of Type II error?
    decreasing it would increase probability Type II error cuz its greater probability of assuming H0 is true when it should be rejected
  6. how to reduce probability of a Type II?
    • two ways:
    • 1. increase alpha,
    • 2. increase sample size (increases power of test)
  7. why might fail to reject null hypothesis?
    • 1. H0 is really true
    • 2. H0 is false, but not enough data to reject H0
  8. assumptions of a t-test for two independent groups?
    assume they are independent of each other, following a normal distribution, means for two groups can be different but variances are equal
  9. degrees of freedom for Poisson and normal?
    Poisson have 1 parameter, means df= k-2

    normal have 2 parameters, means df = k-3
  10. why small value pose a problem for chi-square?
    small values poses a problem cuz cause the chi-square to be artificially high
Card Set