show all formulas and all work - all work is by hand

1. Explain when you would want to use a dependent samples t test instead of an independent samples t test - be specific and complete.

2. You are interested in the effects of low dose alcohol on driving ability. In your lab you have a driving simulator with which you can test subjects and record the number of errors (i.e. driving mistakes) in a test session. You have only 10 subjects (aged 21 or older) that you can test, so you decide to test them in both the placebo condition (no alcohol, just orange juice) and the alcohol condition (1.5 oz of vodka in a glass of orange juice). You decide to counterbalance the test conditions, so you randomly assign 5 subjects to be tested in the alcohol condition first and 5 subjects in the placebo condition first, week one. Then one week later you test the subjects again, but in the opposite treatment group. The data (number of driving errors) is shown below. All subjects are told not to consume any alcohol for 24 hours prior to being tested.

(a) What is the appropriate statistical test for analyzing the data? Why?

(b) State the independent and dependent variables, H0 and H1 hypotheses, level, df, and decision rule.

(c) Perform the analysis (show ALL formulas and ALL computations).

Placebo Alcohol 10 14 9 13 9 13 8 11 7 13 7 10 7 8 6 6 6 7 5 5

(d) What is your decision regarding the H0 and what conclusions can you make from the results of this study?