In: Statistics and Probability
A consumer researcher wanted to know if customers really are
influenced to buy more from sales clerks who smile. To test this,
clerks at eight stores in a large Canadian clothing chain were
given special instructions at the start of a week and then sales
over the week were recorded. Four of the stores were randomly
selected to have the clerks receive instructions to be especially
courteous and to smile a lot. Clerks at four other stores were
simply instructed to be especially courteous. Sales (in thousands
of dollars) for the four stores in the smile condition were 36, 40,
36, and 44; sales for the four stores in the control condition were
40, 31, 27, and 30. Do these results suggest that customers might
buy more if they encounter smiling sales clerks? (Use the .05
level.)
Below are questions you should be asking yourself in preparing for
the test in class (not to be answered here):
Using the five steps of hypothesis testing, think about "why is
this a test for independent means?", "what are the steps involved"
and "what is my goal in this exercise? That is, which distribution
am I trying to construct so I can get the comparison
distribution?"
Question to be answered here:
What is the pooled estimate population variance of the two
populations from which your samples come from?
Questions to answer regarding the question above :
1) In the question above, what is the variance of the distribution of means corresponding to the condition in which the clerks smile? (It may help you to draw the steps using the curves to figure out which variance goes with which distribution). Note that the variance is the same for both distribution of means due to the fact that N is the same for both.
2) In the question above, what is SDifference?
3) In the question above, what is tCutoff? (Answer to the three decimal places on this one)