SOSC 2500 Lecture Notes - Lecture 9: Effect Size, Confidence Interval, Alternative Hypothesis
Document Summary
You state the nh thinking it is false, and then you calculate stats to try to find evi- dence to reject the nh. When you reject the nh, you can accept the alternative (or research) hypothesis) The alternative/research hypothesis stands in opposition to the nh, it asserts something other than the null hypothesis. The alternative hypothesis is more valid if you state it a priori; if you predict it rather than assert it after it happened. Two-tailed: non-directional > there is a difference btwn two groups, but no statement is made about the direction of the difference btwn the two groups. One-tailed: directional > there is a difference btwn two groups and a statement is made about the direction of the difference. So group 1 mean is larger than group 2 mean, or group 2 mean is larger than group 1 mean. Critical t values for two-tailed tests come from appendix b.