27

Question 1

Chi-squared goodness of fit test assumptions? 3•

Answer 1

•SRS of n subjects
•One variable/factor of interest with k mutually exclusive and exhaustive categories/outcomes/levels
•Expected number of outcomes assuming the null hypothesis is true for each level at least 5. aka Expected number of counts is at least 5

Question 2

In chi-squared test, can they be left, right, or both tailed test?

Answer 2

Only right

Question 3

Assumptions of paired t-test? 4•

Answer 3

•two populations (1,2)
•random samples of size n have been collected from each population. The measurement taken from first population is paired with measurement from second population.
•Population of computational interest is the population of differences of these paired observations.
•The distribution of differences is mortally distributed or not extremely skewed with unknown mean and variance.

Question 4

What happens to type II error if you increase sample size?

Answer 4

You decrease the probability of it.