Question 1038866: I am in dire need of help. I'll post the question I'm working on below. Basically, when doing a t-Test: Two-Sample Assuming Unequal Variances in Excel, you need to fill in the ranges used, but I am only given a mean for each test in my homework problem. If you know how to do this or even a faster way to do this problem than typing what feels likes lines of code in excel or on my calculator, please share.
Thanks in advanced.
Problem:
Businesses, particularly those in the food preparation industry such as General Mills, Kellogg, and Betty Crocker, regularly use coupons as a brand allegiance builder to stimulate their retailing. There is uneasiness that the users of paper coupons are different from the users of e-coupons (coupons disseminated by means of the Internet). One survey recorded the age of each person who redeemed a coupon along with the type (either electronic or paper). The sample of 35 e-coupon users had a mean age of 33.6 years with a standard deviation of 10.9, while a similar sample of 25 traditional paper-coupon clippers had a mean age of 39.5 with a standard deviation of 4.8. Assume the population standard deviations are not the same.
1.
Find the degrees of freedom for unequal variance test.
2.
State the decision rule for .01 significance level:
3.
Compute the value of the test statistic:
4.
Test the hypothesis of no difference in the mean ages of the two groups of coupon clients. Use the 0.01 significance level.
Answer by Theo(13342)   (Show Source):
You can put this solution on YOUR website! i was able to do the t-test using excel.
the first part was to create the data.
you do that using the norminv(rand(),mean,standard deviation) function in excel.
it is helpful to see what the actual mean and standard deviation was and redoing the calculations until you get a mean and standard deviation that you are satisfied is close enough to the desired mean and standard deviaton.
this might take several tries, but you will eventually get something in the range you want.
for sample 1, you would create the function =norminv(rand(),33.6,10.9) in one of the cells and then copy that cell another 34 times so you have a total of 35 cells with that function in them in the same column.
for sample 2, you would create the function =norming(rand(),39.5,4.8) in one of the cells and then copy that cel another 24 times so you have a total of 25 cells with that function in them in the same column.
above the two columns (leave enough rows above to put these additional functions in), you would place the following functions in one of the cells on top of each column and in another one of the cells on top of each column.
the functions are:
=average(enter range of cells that contain sample 1 data)
= stdevp(enter range of cells that contain sample 2 data)
it is important that you recalculate the data until you get data that gives you a mean and standard deviation that are close to what you desire.
it may take several iterations.
you cause the recalculation by entering any blank cell on the worksheet and enterinng any data and then hitting the return.
this will cause the whole worksheet to recalculate.
once you get the data you want, it is important to do a copy and a special past for values only on that data so it will remain the same forever.
i did that and this is the data set i came up with after several iterations.
i then ran the datanalysis toolpack in the data tab and executed a t-test with unequal variances.
that test form looked like this.
i then got the results in a separate worksheet that looked like this after i expanded the columns to fit the data.
the results show that the t-score is equal to -3.51 with 49 degrees of freedom and that the probability of getting a z-score less than that is .00095.
the results showed that the critical t-score cutoff point for acceptance that the means are the same was 2.68.
this indicates that the difference between the means was statistically significant and that the likelihood of getting two samples of such different means was not due to random chance.
here's another calculator that confirmed that the means are statistically different.
http://www.evanmiller.org/ab-testing/t-test.html
here's a degrees of freedom calculator for t-test.
it says the degrees of freedom was 49.something which agrees with excel calculations.
http://web.utk.edu/~cwiek/TwoSampleDoF
here's some online tutorials that might help you.
first one is tutorial to create random data with normal distribution.
http://www.mbaexcel.com/excel/how-to-create-a-normally-distributed-set-of-random-numbers-in-excel/
second one is to use the t-test function within the analysis toolpak of excel.
http://www.excel-easy.com/examples/t-test.html
you have to make sure the data analysis toolpak is installed.
then you go to the data tab and execute data analysis and then choose the test you want to perform.
there are also videos on line that help you go through it.
|
|
|
