Reconsider the previous exercises about reaction times and time of day. The statistic used was the mean of the difference in reaction times (morning ˆ’ night). Explain how the following things would change (or not) if we used the mean of the (night ˆ’ morning) as the statistic when the test was conducted?
a. Observed statistic
b. Alternative hypothesis
c. p-value
d. Confidence interval

Data from previous exercise

Student researchers wanted to see whether a person€™s reaction time is affected by the time of day. They had a computer program that would display a red circle. When it changed to yellow, the subjects were to click the mouse as fast as they could and the computer would calculate the reaction time. The researchers had their subjects perform this test at 10:00 am and again at 10:00 pm, recording the reaction time for each student at each time. They randomly determined which time of day was done first for each of their subjects. Here are their hypotheses.
Null: There is no association between time of day and reaction times (μ_d = 0).
Alternative: There is an association between time of day and reaction times (μ_d ‰  0).
The students found the following results for reaction times in seconds:

Transcribed Image Text:

Morning Night Difference Mean 0.244 0.306 -0.062 Standard deviation 0.052 0.084 0.082