Question 1175402: The following reaction times (measured to the nearest second) were recorded for 8 drivers under two different conditions, without consuming alcohol and after consuming a controlled amount of alcohol.
Driver 1 2 3 4 5 6 7 8
Without Alcohol 3 4 3 3 6 5 3 5
With Alcohol 3 5 4 6 7 4 4 7
(a) Test, at 1% level of significance, whether there is evidence that the alcohol increases the reaction time. (12 marks)
(b) Do you think it is necessary to use the same driver for two different conditions above? Explain your reasoning. (3 marks)
Answer by Boreal(15235)   (Show Source):
You can put this solution on YOUR website! This is a paired t-test with df=7
Ho: d=0
Ha: d ne 0
alpha is 3.499 p{reject Ho|Ho true}
calculations
differences are 0,1,1,3,1,-1,1,2
mean is 1
sd is 1.20
t=(x bar-0)/s/sqrt(n)
=1*sqrt(8)/1.20
=2.35
fail to reject Ho at the 1% level. p-value is 0.050.
Insufficient evidence
Using the same driver eliminates differences in metabolism and reaction time, and is a far better technique here.
|
|
|
