Part of a usability study to assess the usability of voting machines included a measure of the time on task (TOT) of voters casting ballots. The data are for the same ballot cast on two different voting machines at the same location.
A few background items:
* The voters (participants/users) are a homogeneous group. * Voters were randomly assigned to theo vting machines. * Due to items 1 and 2 above, assume the two groups of voters (one group using the DRE voting machine, and the other using the OptiScan voting machine) have equal variances. * We have no information to indicate that one voting machine will be faster than the other.
Your job will be to perform a “t” test on these data and draw whatever conclusions you believe you can get from the data. If you need a refresher of the “t” test, read the “t-test description.pdf” document. If you need more information, check your statistics book, or use the Internet to find web sites such as http://www.graphpad.com/quickcalcs/ttest1.cfm. (Excel has a “t” test function although it may not be currently installed in your version; you would then add it in.)
Question 1 – What is the null hypothesis in this evaluation? (Discussed in class, but easy found on the Internet)
The voting machine brand does not distinguish difference in the voting time.
Question 2 – Which “t” test should be used – paired, unpaired/equal variance, unpaired/unequal variance?
Unpaired/equal variance should be used because each group had one data point for every occurrence.
Question 3 – Should a one-tail, or two-tailed test be used, and why?
Two tail test should be used there are two groups or products to compare.
Question 4 – Is the t value significant at the 0.05 level, and why?
No it is not significant because the p value is smaller than the alpha value .05.
Question 5 – Based on (1) the above analysis what, if...