Posted: May 3rd, 2023
Statistic
Collegedata.com has a review of Rutgers University — New Brunswick in the following website
https://www.collegedata.com/cs/data/college/college_pg01_tmpl.jhtml?schoolId=1441
Some interesting facts you may find there (for college students)
· The freshman retention rate, an indicator of student satisfaction, is 93 percent.
· 43 percent of the students live in college-owned, -operated, or -affiliated housing
· 53.3 percent of undergraduates receive some kind of need-based financial aid
· 41 percent of enrolled freshman have SAT math score between 600 and 700
· 14.5 percent of all students from U.S are Hispanic/Latino
Step 1: Choose one of these five percentages that is most interesting to you to set up hypothesis testing. You may choose to have one-tailed or two-tailed test, just explain the reason why you choose it. For example, if you suspect that the percent of students who received some kind of need-based financial aid is too high, you may have your hypothesis as H0: p=0.533, Ha: p<0.533
You also need to test the hypothesis that no difference exists between the proportion of female who received some kind of need based financial aid and proportion of male who received some kind of need based financial aid, i.e. H0: p1-p2=0. Again, you may choose to have one-tailed or two-tailed test, just explain the reason why you choose it.
Step 2: Go collect the data
For example: at the door at the dining hall, you may wish to interview the students randomly and ask whether they received some kind of need-based financial aid.
DO: Take random samples. Be careful of your target population. For example for the first fact, you should only get data from freshman. All the percentages here are for college students (so don’t interview graduate students). Take independent samples. Let your sample size be large. (at least 15 success and failures for both male and female samples)
DON’T: Don’t take convenient samples. For example, don’t just ask your buddies the question, or just knock the door of your neighbors in the dorm. DON’T take dependent samples. For example, don’t ask a guy the question, and then ask him whether his girlfriend…
Step 3: Do the hypotheses testing and show your conclusion whether there is sufficient evidence to support collegedata.com claims, and whether there is difference between the proportions in male and female students.
Submission format and grading rubric
· (10 points) Description of your data collection method
· (10 points) Raw data you collect (including observation number, gender, 1/0 to your question. You may also want to have the freshman, or U.S residency status if you are interested in the first, fourth or last fact)
· (10 points) Description of the parameter in the project
· (10 points for each hypothesis) Description of your hypotheses tests and why you set up them like that
· (20 points for each hypothesis) Hypothesis testing results (including significance level, test statistic, rejection region, p-value and conclusion)
· 10 points for the overall structure of the report