Question 538:

Asked on December 9, 2008

Tags: ANOVA , Paired t-test

1

Answer:

No answer provided yet.

You would use an ANOVA when you want to compare the means for more than 2 groups. For example, if you wanted to see if there was a significant difference in the average time between four cashiers to ring-up 15 items you could use an ANOVA. If you were comparing only 2 groups you could use the independent samples t-test.

A paired t-test is used when you have matched data in two conditions. For example, if you wanted to see if there was a significant difference in the average number of correct responses to a math test after sleep deprivation. You could give the test to 30 students after a normal nights sleep (8 hours) then again three days later with only 1 hour of sleep.  Since you have the same participants in both samples you would use the paired t-test, since the subjects are NOT independent. The advantage to the paired-t test over the independent sample t-test is that you essentially eliminate the between subjects variability when you use the same people. This is why it is often called a within-subjects design. This type of design has higher power, meaning you can detect smaller differences between conditions. The disadvantage is often carry-over effects from the subjects from the first condition to the second condition, as such measures need to be taken to mitigate this nuisance effect--like using different math questions.