Question

A study in 2007 involved 118 college students who volunteered to fill out a survey in exchange for $5. After completing the survey, the students were asked if they would like to donate any of their $5 participation fee to a real, legitimate charity that addressed global hunger. A randomization led to half of the students (group 1) being shown a picture and description of a young girl who had been helped by the charity, and the other half of the students (group 2) being given statistical information about the effectiveness of the charity (but nothing about any individual child and no pictures). The sociologists wanted to test whether there would be a stronger response to the plight of an individual, pictured child than to statistics about all the children. The data that was collected was the amount of money the college students each donated from their $5.

6) Write the null and alternate hypotheses about the mean amounts donated.

7) We have no idea if the population or sample data in this study are approximately symmetric. Why can we continue with t- distribution procedures?

8) The sample mean from group 1 was $2.12, and the sample standard deviation was $2.13. The sample mean from group 2 was $1.21, and the sample standard deviation was $1.67. What is the difference between the sample means?

9) Use 2-SampTInt to calculate a 90% confidence interval. Is there statistical evidence that the students gave more money whenpresented with an individual’s information rather than just being given general information? Why or why not? (Use the STATSoption. Choose No for Pooled.)

10) WITHOUT calculating it using a T-test, is the P-value less than 0.05? Why? (Use the results from the confidence interval as a guideline).

Answer 1

Ans: There is statistical evidence that the students gave more money when presented with an individual’s information rather than just being given general information at 0.10 level of significance because the confidence interval does not include value zero.

10) WITHOUT calculating it using a T-test, is the P-value less than 0.05? Why? (Use the results from the confidence interval as a guideline).

Ans: Yes, the p-value is less than 0.05 because of the 90% confidence interval does not inlude zero.