Presentation is loading. Please wait.

Presentation is loading. Please wait.

Predicting Academic Performance of University Students

Similar presentations


Presentation on theme: "Predicting Academic Performance of University Students"— Presentation transcript:

1 Predicting Academic Performance of University Students
BIT 5534: Applied Business Analytics & Business Intelligence Team 4: Michael Cerney, Chris Kopinski, and Chris Stewart

2 Agenda 1. Business Problem 2. Data Understanding 3. Data Preparation
4. Modeling (Ordinary Least Squares + Decision Tree) 5. Results 6. Conclusion

3 Business Problem Topic: University student attrition and retention
Business Problem: Identify characteristics/attributes of success or failure for University students Measure of Success: GPA Performance What’s to be gained? Universities can have more successful student recruitment and selection Universities create programs that encourage success of current students Miles from home? Part-time work hours? Accommodations? Business Problem Data Understanding Data Preparation Modeling Results Conclusion

4 Attribute Description
Data Understanding Sample of Variable Dictionary Dataset acquired from JMP textbook website Attributes are student-centric: Ex. GPA, College, Accommodations, etc. GPA identified as dependent variable Threshold for academic success based on GPA but not intentionally defined Attribute Name Variable Type Attribute Description GPA Continuous GPA while attending university Miles from Home Distance from campus Accomodations_Dorm Student lives in dorm Attends Office Hours_Never Student never attends office hours College_Business Student majors in Business Class_Freshmen Student is a freshmen Business Problem Data Understanding Data Preparation Modeling Results Conclusion

5 Data Preparation Data Consolidation Data Cleaning Data Transformation
Data Selection: Academic variables of interest identified Data Cleaning Missing Values Report: No reported values missing Outlier Detection Report: No reported outliers Data Transformation Dummy variable creation: College Attends Office Hours Accommodations Class Data Reduction Removed variable Return Removed 23 records that included the Return attribute Business Problem Data Understanding Data Preparation Modeling Results Conclusion

6 Modeling – Ordinary Least Squares
OLS Model Independent Variable Estimates Statistically significant and positively correlated with GPA based on (p < 0.05): College_Business, College_Sciences, College_Engineering Statistically significant and negatively correlated with GPA (p < 0.05): Accomodations_Off-campus Business Problem Data Understanding Data Preparation Modeling Results Conclusion

7 Modeling – Decision Tree
Decision Tree contains 7 splits and records: First split = yes/no for Business School Business school in general has higher GPA (2.56 vs 2.44) Second split = Business School + yes/no on off campus accomodations Off campus has a higher GPA (2.62 vs 2.51) 1 2 Business Problem Data Understanding Data Preparation Modeling Results Conclusion

8 Results – K Fold Cross-Validation
K-Fold Cross-Validation Technique Dataset was partitioned into five subsets (215 records each) Each training set contain 80% of the dataset, formed into a unique combination of subsets Each validation set contain 20% of the dataset (1 unique subset) Training Sample Subsets Validation Subset 1 A,B,C,D E 2 B,C,D,E A 3 C,D,E,A B 4 D,E,A,B C 5 E,A,B,C D Business Problem Data Understanding Data Preparation Modeling Results Conclusion

9 Results OLS Model Results Decision Tree Model Results
Confirmed the statistically significant variables (Business, Engineering, Science) Living off campus showed statistical significance for higher GPA scores An average (P<0.1) for 5 out of 5 training sets identified that Attends Office Hours Never, displayed a high correlation to lower GPA scores Decision Tree Model Results Students that were enrolled into the college of business and science, and that lived off campus, maintain a higher GPA compared to the students that lived on campus Students that were enrolled into the college of engineering, and lived on campus, performed better than the students that lived off campus Business Problem Data Understanding Data Preparation Modeling Results Conclusion

10 Conclusion Encourage on-campus students to attend office hours
Obtain new perspectives by interviewing students and staff to further investigate low GPA scores Promote mentorship program Increase GPA scores and lower attrition rates Business Problem Data Understanding Data Preparation Modeling Results Conclusion


Download ppt "Predicting Academic Performance of University Students"

Similar presentations


Ads by Google