if the test statistic is higher than the upper critical value that means we must reject the null, there was significance.

t test =

• sample avg score x null hypothesis (no diff) overestimated standard error of the mean difference

after testing you want to figure out the 95% confidence interval

know the SD?

• yes - use z
• no - use t

-

• one sample t focuses on the scores
• paired sample focuses on the difference

independent scores t-test - we have two samples, comparing two samples to each other

• four assumptions
  • normality
  • random sampling
  • independence
  • equal variance

looking for the difference between two groups that are independent of each other with no overlap [comparing sample 1 with sample 2]

satisfied when the larger variance is not greater than two times that of the smaller*