DATA DESCRIPTION

A campus survey is conducted at a large size university. The purpose of the survey is to collect back ground information of students and to understand what factors may explain university student’s grade point average (GPA). a sample of 141 undergraduate students is randomly selected. These selected students answer questions on their personal backgrounds and agree to authorize the survey conductor to retrieve their GPA sat the university and high school/college and achievement test scores (A score). The relevant information is entered into a spreadsheet CampusSurvey.xlsx, where each column represents a variable. These variables include:

1. Obs No: A number functioning as an ID for each randomly selected student in the sample.

2. year: The year of university undergraduate program which each student is at: 1st year, 2nd year, third year, or the Honours year.

3. age: age in years of each student gender: female or male

4. campus: = Yes if living on campus = No if not living on campus

5. major: The major of each student

6. uniGPA: University grade point average

(For example, GPA 3.5- 4.0. refers to 90-100%)

7. hsGPA: high school/college grade point average

8. Ascore: achievement test score. An achievement test is a test taken to improve a

9. student’s credentials for admission to universities.)

10. computer: = Yes if a student owns a laptop or a desktop = No if a student does not own a laptop nor a desktop

11. 11. bgfriend: = Yes if a student has a girl/boy friend = No otherwise

12. skipped: average number of lectures missed per week

13. alcohol: average number of days per week the student drinks alcohol

14. job20: = Yes if a student works no less than 20 hours per week = No otherwise

15. volunteer: = Yes if a student does volunteer work = No otherwise

Task 1 Describing univariate data

Pick up three numerical variables and three categorical variables that you think are important information of students. Write a report to describe them one by one. To describe univariate categorical variables, you need to use appropriate univariate tables or/and charts. To describe univariate numerical variables, you need to use appropriate univariate tables or/and charts as well as appropriate numerical measures to detail distribution of the numerical variables. 

Task 2 Describing bi-variate numerical data

Use an appropriate graphical technique and an appropriate numerical measure to discuss the relationship (if there is any) between the university GPA and each of the numerical variables provided in the dataset. Based on your analysis, which numerical variable is most related to university GPA?

Task 3 Describing bi-variate categorical data

You want to discuss whether there is a relationship between uniGPA and a categorical variable that describes student’s personal background. For example, do students who own a computer tend to have higher university GPA? Do students who work no less than 20 hours per week tend to have lower university GPA? Does living on campus affect university GPA? etc…

Our challenge is that uniGPA is a numerical variable, not a categorical variable. One way to work on two different types of variables is to transform one variable to the type of the other.

You now decide to generate a new categorical variable based on the level of university GPA. Since a GPA score 3.0 and above indicate that the final mark is no lower than 85%, you decide to use the value of 3.0 as a threshold score to generate the new categorical variable.

Enter “High” if a student’s uniGPA is no less than 3.0 and enter “Low” otherwise. Choose a categorical variable from the dataset that you believe may potentially affect student’s university GPA, and write a short report to discuss whether you observe any relationship. Your report should include the following:

I. Present these two categorical variables, the newly generated variable based on uniGPA and the other categorical variable of your choice, together using an appropriate graph.

II. Produce a contingency table of frequencies to present these two categorical variables. Based on the contingency table, calculate relevant joint probabilities, marginal probabilities, and conditional probabilities.

III. Your discussion on relationship between the two categorical variables need to be based on the graph and appropriate probabilities.

Task 4 Inferential analysis – hypothesis testing

In the previous question, you simply use the GPA score 3.0 as a given threshold score to distinguish between students who have high GPAs with students who have low GPAs. One may argue that it is a sensible choice only if the population average uniGPA level is 3.0.

Conduct a hypothesis test to discuss whether the choice of using 3.0 is sensible. The test is performed at 5% level of significance.

file-of-data-campussurvey