2. There are many factors that can affect a student’s GPA, both internal
and external. For our project, we were interested in looking at the
affects of a couple of external factors: coarse load and employment
status
In order to begin our research, we created an online survey, the link
to which we posted on social media
3. In order to gather information
about students at the
University of Houston, we
created a survey on
surveymonkey.com and
posted the link to social
media. We received 65
unique responses to our
survey.
5. We expect:
68% of the data to fall within 2.572 and 3.963
95% of the data to fall within 1.878 and 4.0
99.7% of the data to fall within 1.183 and 4.0
In reality:
89% of the data fell within 2.572 and 3.963
95% of the data fell within 1.878 and 4.0
99.7% of the data fell within 1.183 and 4.0
From these measurements, we can deduce that the data is not normally
distributed.
6. Our first question considered how working a job while in
school affected a student’s grade point average.
7. 0
2
4
6
8
10
12
0 - 2.00 2.00 - 2.5 2.5 - 3.00 3.00 - 3.5 3.5 - 3.75 3.75 - 4.00
NumberOfStudents
GPA
GPA vs. Number of Hours Worked Per Week
40 hours /week 30 - 20 hours/week 15 - 0 hours/week
GPA Range
40 or more
hours /week
30 - 20
hours/week
15 - 0
hours/week
Total by
GPA
0 - 2.00 2 0 1 3
2.00 - 2.5 2 2 0 4
2.5 - 3.00 3 3 5 11
3.00 - 3.5 1 5 11 17
3.5 - 3.75 2 1 8 11
3.75 - 4.00 2 5 12 19
Total by hours 12 16 37
14. Number of Hours Worked per Week GPA
40 or more 2.19 – 3.31
30 2.56 – 4.06
20 2.93 – 3.53
15 3.05 – 3.54
10 – 5 2.65 – 4.01
0 3.37 – 3.69
15. Hypothesis – Students at the University of Houston that work less than 15 hours a
week have a higher GPA than those who work more than 20 hours a week.
Null and alternative hypotheses
H0: (m<15 – m>20) = 0
Ha: (m<15 – m>20) > 0
Test statistic
z = 2.021
Level of significance: 5%
P-value
p = 0.0217
16. Our second question considered how taking different
class loads affected a student’s grade point average.
24. Number of Credit Hours per Semester GPA
18 or more 3.46 – 4.01
15 3.34 – 3.59
12 2.70 – 3.30
9 2.93 – 3.76
6 or less 1.42 – 3.50
25. Hypothesis – Students at the University of Houston that take 15 hours will have a
higher GPA than all other students.
Null and alternative hypotheses
H0: (m15 – m!15) = 0
Ha: (m15 – m!15) > 0
Test Statistic
z = 2.788
Level of significance: 5%
P-value
p = 0.0026
26.
27. In both hypothesis tests, we found evidence to reject the null hypotheses of no
difference between GPAs in favor of our alternative hypotheses, which theorized that
both holding a job and taking differing amounts of course work affected GPA
The confidence intervals for the difference between means are in further support of
those ideas:
(m<15 – m>20) = 0.122 – 0.700
(m15 – m!15) = (-0.707) – (-0.0108)
28. Mendenhall, William, Robert J. Beaver, and Barbara M. Beaver. Introduction to
Probability and Statistics. Pacific Grove, CA: Brooks/Cole, 2013. PDF.
“How does having a job affect a student’s GPA?” SurveyMonkey: Free Online Survey
Software & Questionnaire Tool. Jessica Bradham, 3 March 2015. Web. 18 April 2015.
<https://www.surveymonkey.com/r/6NMTJFT>.
Editor's Notes
Our team did a survey on how holding a job along with their class load can affect their Grade Point Average.
There are many factors that can affect a student’s GPA, we focused on how the amount of hours on their course load and the amount of hours working. We created an online survey and received 65 responses.
The survey was created on Survey Monkey, asking these 6 questions to gather our data.
Using the data retrieved, we were able to find out the information needed
According to the empirical rule, the data revealed 89% of students’ GPA that fell within 2.572 and 3.963 and 95% that fell within 1.878 and 4.0. From these measurements, the data was not normally distributed.
Our first question considered how working a job (measured by hours worked per week) while in school affected a student’s grade point average.
This graph represents the average GPA for students who work 40, 30, 20, 15 hours per week or do not work. We can see that the average GPA is lower for students who work full time than for those who work part time
This graph represents a middle value for GPA. We can see that the GPA is higher for students who work less than 10 hours per week
The range represents the difference between the highest and lowest GPA. As we can see, the most significant change is for 40 hours and 10 hours per week
The lower quartile represents the median of the lower half of the data set. As we can see, the lower quartile is almost identical to the median graph.
The upper quartile represents the median of the upper half of the data set. In our example, the upper median is significantly higher for students working 30 hours per week and 10 or less hours per week
The standard deviation measures how far typical values tend to be from the mean. We can see that data representing 40 hours per week and 10 hours per week are 1 standard deviation from the mean. The rest of the data is less than one standard deviation from the mean, so the values in a dataset are pretty tightly bunched together
The confidence intervals show an estimated range into which the mean of the population will likely fall. Here, we have calculated 95% confidence intervals of the mean of the GPA for each category of students based on how much they work per week. We can begin to infer from the intervals that those students who did not work at all likely have the higher GPAs when compared to students who work at least a few hours per week.
Our hypothesis is that students at the University of Houston who work less than 15 hours a week will have a higher average GPA than students who work more than 20 hours per week. In opposition, our null hypothesis states that the mean GPAs of students who worked 15 hours or less and that of students who work 20 hours or more are equal. We set our level of significance at 5%, and calculated a test statistic of 2.021 from the data we collected. Using Table 3 in Appendix 1 of the textbook, we found a p-value of 0.0217, therefore, we reject the null hypothesis. This means that there is statistically significant evidence that our alternative hypothesis is correct.
Our second question considered how taking different class loads (measured by credit hours per semester) affected a student’s grade point average.
This graph represents the average GPA for students who take 18, 15, 12, 9 or 6 credit hours per semester. We can see that surprisingly the average GPA is higher for students who take the max amount of credit hours.
This graph represents a middle value for GPA. Again, we can see that the GPA is higher for students who take 18 credit hours per semester
The range represents the difference between the highest and lowest GPA. As we can see, the most significant change is for 12 credit hours and 6 credit hours per week
The lower quartile represents the median of the lower half of the data set. As we can see, the lower quartile is almost identical to the median graph.
The upper quartile represents the median of the upper half of the data set. In our example, the upper median is significantly higher for students who take 18 credit hours per semester and 9 credit hours per semester
The standard deviation measures how far typical values tend to be from the mean. We can see that data representing 6 credit hours are 1 standard deviation from the mean. The rest of the data is less than one standard deviation from the mean, so the values in a dataset are pretty tightly bunched together
Here, we have constructed 95% confidence intervals of the mean of the GPA for each category of students based on how many credit hours they take per semester. These results show a slightly more interesting idea that those for working hours. We can infer from the results that students who take 18 credit hours or more every semester have the highest range of mean values for the population, despite having the most school work every semester.
Our hypothesis claims that those students who take 15 hours per semester have a higher average GPA than all other students. In opposition, the null hypothesis states that there is no difference in the means of students who take 15 credit hours and those who take different amounts. We set our significance level at 5%, and calculated a test statistic of 2.788. From this, and again using Table 3 in Appendix 1, we find a p-value of 0.0026. Therefore, we reject the null hypothesis. This means there is statistically significant evidence that our alternative hypothesis is correct. It is worth noting that this does not mean that students who take 15 credit hours per semester have the highest GPAs of any other group of students because they have a higher GPA than all other students combined. As you can see in the confidence intervals and the graph of all of the groups, students who took 18 credit hours actually had the highest GPAs of any other group.
The conclusions we drew from the data are that both holding a job and taking on different course loads are factors that will affect a student’s GPA. In both hypothesis tests, we found evidence to reject the null hypotheses of no difference between GPAs in favor of our original hypotheses that these factors made a difference. The confidence intervals for the differences in means shown here further support our ideas, as they do not include zero, and therefore support the likelihood of a difference existing.
Over the course of our project, we used the textbook as a reference, and hosted our survey at surveymonkey.com