Quantitative Methods II
Issue: Student success in their first year is a strong predictor of subsequent degree success.Students experiencing difficulty in first year have a high likelihood to not return for their second year.Interventions to assist at risk students must occur early in the first year if they are to be effective.Ideally, they should start in the first two months of the fall term.This is before the first midterm results are available.Can we identify who will be unsuccessful (or successful) before classes even start?
Questions:What is success for a new student?Getting a GPA above 2.00?Keeping your entrance scholarship?Performing above average compared to similar peers?
What data is available?We will limit attention to data that is captured in Banner.In recent years, we have received electronic transcripts for all Nova Scotia high school applicants raising the possibility of knowing student performance, by course, as far back as grade 10.However, for this study, we will look at admissions data from 2015, before these transcripts were available.
For applicants from Canadian high schools, the high school average (out of 100) was recorded at the time of the admission decision.The average was based upon the 5 courses used for the admission decision.If the admission decision was made in the fall while the student was in grade 12, the average was based upon final grade 11 grades.If the decision was made later, it could include some final or mid-year grade 12 course grades.
Exclusion: High school average data for international students is inconsistent, so we will restrict analyses to Canadian students only.Many new students are transferring from colleges or universities.This is a very diverse student population for whom high school data may not be available or may not be a good predictor of success.We will not include these applicants in this study.Identifying potentially unsuccessful international or transfer students should be addressed in separate studies.
Other admission data includes the date of application and the date of the admission decision, the name of the high school and the province in which it is located, and the student’s program of study.Note that admission requirements vary among Arts, Business, Science and Engineering, so the courses included in the high school average will be different.Similarly, the recommended mix of first year courses varies by program.
The extract occurred in January 2016 and included only those students admitted for September 2015.The fall GPA was recorded for each student.
Exclusion: Some new students drop out during the fall term.Some register but never actually attend.These students will show with no earned credits and a GPA of zero.We will consider them as a distinct population of students that should not be included in this study.We will delete all students with a fall GPA of zero.This action deletes students who were active in the fall but failed all courses.As such, some valid records have been deleted, but there was no way to identify these students separately from those who simply disappeared.
In addition to the above “data cleaning”, several variables have been recoded.It is impossible to look at individual high schools, but this information was used to distinguish between students from the Halifax region and those from elsewhere in Nova Scotia.Sample sizes from many provinces outside Nova Scotia are small, so they have all been grouped together as Out of Province (OOP).Similarly, because of the small sample sizes, Engineering students have been merged with Science students and Environmental Studies with Arts.
Although significant data preparation has been done, you will need to do more data preparation to complete assignments 3 and 4.Some will be necessary to answer the questions posed and some will be needed to conform to the way Excel does data analysis