Information item# 3 -Submitted by the Strategic Planning Advisory Committee (2015/2016- Dec)

 

Determinant of the SixYear Graduation Rate:
Findings from a Multi-Year Study
 by Yu, Fen and Powell, Rick

Result is based on FTIC students for 2003 – 2013 at UNF using a Logistic Regression Model.

 

High School Academic Variables

Beta Coefficient

P-­‐value

Odds Ratio

 

 

 

 

High School GPA

-.017

.548

.983

AP Courses

.066

.010 *

1.068

Honors Courses

-.051

.139

.951

Math Courses

-.001

.988

.999

ACT/SAT

-.182

.000 *

.834

 

 

 

 

College Variables

 

 

 

 

 

 

 

College GPA 3 or above

1.777

.000 *

5.914

Averaged 12 Hours

.428

.000 *

1.535

Honors Program

.480

.000 *

1.616

Lived on campus

.023

.624

1.023

Residency (FL or not)

.580

.000 *

1.786

 

 

 

 

Financial Aid Variables

 

 

 

 

 

 

 

Need based Scholarship

.295

.000 *

1.344

Non-Need based Scholarship

.260

.000 *

1.297

Work Study

.402

.000 *

1.495

All the * ‘s mean the variables are significant in the result.


Interpretation of the result:

For  cohort  tracking  2003  to  2013  for  six- year  graduation,  a  logistic  prediction   model  identifies  high  school  academic  variables,  such  as  GPA,  honors  or  math  courses  are  not  at  all  significant  but  ACT/SAT  score  is  significant  with  a  negative coefficient,  meaning  high  school  credentials  are  not  impacting  much  the  graduation rates  except  AP  courses  taken  at  school.  The question is now what does.    The  data shows  that  the  college  variables,  college  GPA  3.0  or  above,  average  12  or  more  credit hours  taken,  Honors  program,  FL  residency  status,  scholarship  given  (need  or  non- need),  and  Work  Study  are  all  highly  significant.  Whether  the  student  lived  on  campus  was  not  a  significant  variable.