Health Sciences 3801A/B Lecture Notes - Lecture 5: Phi Coefficient, Standard Deviation, Randomness
Document Summary
Proceed in the same way as the univariate chi-square analyses. You subtract them to get the mean score. If a person has a higher mean score at time 2 than time 1, you can say that they increased in a positive direction. Analyzing continuous variables - you can use a t-test test. If it is nominal or ordinal, you need a new statistic that doesn"t require the mean or standard deviation & will allows us to look at change within nominal or ordinal variables. Cramer"s v can be used for two nominal variables. Randomness would be seen in the chi-square calculations. Proceeds in the same way as the previous chi-square calculations. If the chi-square obtained is greater than the critical value then the variables are correlated. We can calculate this to determine which cell deviates from the expectation. Scores that are greater than 2 or less than -2 are considered to be statistically significant.