Question 135008
A manager hopes to improve the sales figures for 10 employees by sending them to a motivational seminar.  To test the effectiveness of this seminar, the manager compares the sales figures for each salesperson for the month immediately prior to the seminar and the month immediately following it.  
Employee	before	after
Homer	19	12
Bart	6	18
Lisa	9	12
Marge	5	7
Maggie	9	14
Ned	8	17
Milhouse11	9
Willie	11	17
Stu	11	12
Gill	7	7

Run a statistical hypothesis test and determine if the seminar significantly improved sales figures for the company.  
--------------------------------------
These are dependent sample pairs because each pair related to a given
person's performance before and after the seminar.
-----------------------------------------------------
Enter your data;
Calculate the difference for each pair (before-after):
Run a t-Test on mean of the differences.
---------------
Comment: If there is improvement before minus after should be negative.
---------------
Ho: mu(differences) = 0
Ha: mu(differences) < 0 (Claim)
------------------
d-bar(mean of the differences) = -2.9 ;s(diff) = 5.4661
---------------------
Critical Value for one-tail t-test with df=9 and alpha = 5%: t = -1.833
----------------------
Test statistic: t(-2.9) = (-2.9-0)/[5.4661/(sqrt(10)] = -1.6777
----------------------
Conclusion: Since test statistic is not in the rejection interval,
Fail to reject Ho; the seminar made no statistical difference in sales.
----------------------
Cheers,
Stan H.
-----------------