TO:            Faculty Senate, University of Oklahoma


FROM:      Faculty Senate Task Force on Student Graduation and Retention:

Roger Harrison, Task Force Chair, Chemical Engineering and

Materials Science

                        Jack Kasulis, Marketing

                        Mike McInerney, Botany and Microbiology


Date:          October 25, 2002


Subject:      Report on Findings about Undergraduate Student Graduation and




President Boren has made the improvement of OU’s graduation rate as one of his highest priorities.  He has announced the goal of increasing the graduation rate at OU to at least 60%, the median of the universities of the Big12/Big 10 within four years.  Several groups have been formed to study this issue, including this Faculty Senate task force that was organized in October, 2001.  This task force was formed with the objective of evaluating as much information as possible about the graduation and retention of OU undergraduate students.  This report will present and discuss our major findings and make recommendations for future action to improve graduation and retention.


The primary retention statistic that is used for comparing universities is the percentage of students who graduate within six years.  The most recently available data on the retention rates of first-time students at OU is summarized in Table 1, which shows the data for students who started at OU in years from 1990 to 2000.  The six-year graduation rate has risen fairly steadily from 44.0% for students who began in 1990 to 51.1% for students who began in 1995.  Compared to the students who started in 1995, the students who started in 2000 have a first-year retention rate 4% higher, which means that the six-year retention rate for the students who started in 2000 should be at least 55% based on the past trends shown in Table 1.


Complete Withdrawal Survey Results


One of the first sets of data that became available to us was a complete withdrawal survey done by Sophia Morren, Assessment Coordinator.  The most recent of these surveys is shown in Table 2 for the entire campus for the spring 2002 semester.  The reasons for withdrawal ranked in order of highest to lowest percentage are (1) personal, (2) academic, (3) work/employment, (4) financial, (5) other, (6) health, and (7) commuting/transportation.  Some of the reasons under the personal group are strongly related to the academic group, such as “cannot concentrate/haven’t been able to study,” “change of major/don’t know what to major in,” and “boredom.”  Also, “family problems” (22%) in the personal group could indicate that the family is unhappy with the student’s academic performance and does not wish to continue to support the student financially.  When the three personal reasons that are strongly related to academics are included under academic, then the total for the academic group of reasons increases to 65% for spring 2002, indicating that academic problems appear to be the major reason for students leaving the university.  Work-related problems (36%) are another major reason that students give for leaving the university.    


A problem with interpreting this survey is that students do not prioritize their reasons for withdrawing and give more than one reason for withdrawing, with the result that the total of the percentages given for withdrawing are much greater than 100% (for example, 196% in spring 2002).  It is our understanding that future withdrawal surveys will attempt to rank the reasons for withdrawing.


Analysis of Transcripts of Students Who Did and Did Not Graduate


In order to make any recommendations about what the university should do relating to academics to improve student retention, the committee felt that more information on the academic records of those who did not graduate from OU was necessary.  We felt it was necessary to compare the records of the drop-out group to the records of those who graduated within six years.  The most recent group of students who graduated within six years is the group that started in 1995.  We analyzed the student advisory records of 100 randomly selected students who started in 1995 and had not yet graduated compared to  the same records of 100 randomly selected students who started in 1995 and did graduate within six years.  (We provided two sets of random numbers to Cheryl Jorgenson, and she worked with the Academic Records Office to obtain copies of the student advisory records.  Each record showed only a number and did not show the name of the student.) 


The results comparing the records of 100 students who did not graduate to 100 students who did graduate are shown in Table 3.  To check the validity of using sample sizes of 100 randomly selected students for comparison, the retention rates of the 1995 cohort were calculated based on the random sample of 100 students who did not graduate, shown as a note to Table 3.  The one- and two-year retention rates were within a few percent of those found for the 1995 cohort as a whole (Table 1), which gives credibility to the comparison methods used.


For the comparison of the students who did and did not graduate (Table 3), the largest difference between the academic records of the two groups was between the OU cumulative GPA’s and the combined cumulative GPA’s and also between the percentage on academic probation.  The average GPA’s of the students who did not graduate were much lower than for those who did graduate (for example, 1.97 and 3.24 OU cumulative GPA’s for those who did not and did graduate, respectively), indicating that the former group was struggling academically across the board.  Forty-one percent of the group who did not graduate was on academic probation at one time, compared to only 6% for the group who did graduate.  The ACT composite, math, and English scores were slightly lower for the group that did not graduate (less than 1 point difference between the two groups).  Regarding remedial math, the group that did not graduate had a much lower remedial math completion rate than those who did graduate (42% vs. 72%).  Another interesting finding is that men had a higher rate of dropping out than women (females made up 40% of the group that did not graduate and 56% of the group that did graduate).


For the group of 100 students who did not graduate, data are also shown in Table 3 about how long they spent at OU before dropping out.  During any given year, the highest number (44) dropped out after completing one year but before completing two years.  A total of 83 students dropped out before completing three years at OU. 


We feel that this new data is the most significant of any we have analyzed regarding retention and graduation.  Even though students give a variety of reasons why they are withdrawing from OU (see Table 2), all of them on average have one common denominator:  they are poor students, as indicated by the fact that their average OU cumulative GPA is 1.97 (see Table 3).  The data on when the students dropped out of OU indicates that in order to have the greatest impact, any new programs to improve student retention should focus on freshmen and sophomores who are doing poorly academically.     


Statewide Conference on Student Retention and Graduation, November, 2001


One of us (Roger Harrison) attended the Statewide Conference on Student Retention and Graduation sponsored by the Oklahoma State Regents on November 2, 2001 at Rose State College.  The most useful information gained at this conference was from the talk by the keynote speaker, Dr. Clifford Adelman of the U.S. Department of Education.  His findings are based on longitudinal studies of students starting in the 10th grade and continuing until age 30.  These studies indicate that the academic intensity and quality of one’s high school curriculum is much more strongly related to college graduation than either high school GPA or test scores.  Table 4 shows the high school units at the highest level of their academic intensity that Dr. Adelman used to quantify academic intensity.  He modified the curriculum intensity variable for quality, adding gradations for the number of advanced placement courses, highest level of mathematics reached in high school, and subtracting for any case where mathematics course work was largely remedial.


Dr. Adelman also presented interesting information about the influence on degree completion of the highest level of math taken in high school.  He presented the “math ladder” that shows the college degree completion rate as a function of the highest level of math studied in high school (Table 5).  The math ladder shows that students can greatly increase their chances of graduating from college by taking math courses in high school beyond Algebra II; Dr. Adelman believes the reason for this is that many university courses other than just math courses now require math skills for success.  At OU, for example, two science courses must be taken to satisfy the general education requirements.  Many science courses require skills in math for successful completion.  


In the handout that accompanied Dr. Adelman’s talk, two of his major conclusions are the following: 


·        When the academic intensity and quality of one’s high school curriculum is such a dominant determinant of degree completion, and both test scores and (especially) high school grade point average or class rank are so much weaker contributors to attainment, college admission formulas that emphasize test scores and (especially) high school grade point average or class rank are likely to result in lower degree completion rates.


·        The type and amount of remediation matters in relation to degree completion.  Increasingly, state and local policy seeks to constrict—if not eliminate—the amount of remedial work that takes place in 4-year colleges.  But there is a class of students whose deficiencies in preparation are minor and can be remediated quickly without excessive damage to degree completion rates.


OU’s Data on the Influence of High School GPA and ACT Score on Freshmen Success


OU has collected data for three freshman classes, 1999-2001, that show the first semester GPA at OU as a function of high school GPA and composite ACT score (Table 6, provided by Dr. Myrna Carney, Assistant Dean of University College).  These data show that freshmen with high school GPA’s of 3.50 and above do relatively well at OU, regardless of their ACT score.  As indicated in Table 6, approximately 93% of the students with a GPA of 3.50 and above in high school (most of the shaded area) had above a 2.0 GPA in their first semester at OU.  These data also indicate that a student’s high school GPA is a stronger predictor of success in the freshman year than his/her ACT score.  As an example, 85% (on average) of students with ACT scores of only 18 or 19 had an a GPA of above 2.0 during their first semester at OU.  


This set of data could be used as a basis to expand the Gateway to College Learning courses to include more at-risk students.  For example, it would be reasonable to require the Gateway course for all entering freshmen with a high school GPA below 3.00 (846 students during 1999-2001, or an average of 282 students per year).


The effectiveness of the Gateway program is shown in Figures 1 and 2 (provided by Dr. Myrna Carney).  For each group of students (grouped by high school GPA), the Gateway program had a positive effect on either the percent returning the second semester or the percent above 2.0 GPA the first semester.  The effect of the Gateway program was more pronounced for students with less than a 3.50 high school GPA than for those with a 3.50 GPA and higher in high school.


Conclusions and Recommendations


1.      The data in Table 3 lead us to conclude that there must be early and aggressive intervention and then continuous monitoring of students who perform poorly at OU, starting as early as the completion of their first semester here.  This intervention should focus primarily on freshmen and sophomores, since the majority of the students who drop out do so before completing two years at OU.  Counselors to these students must be prepared to help them solve a variety of problems that include financial, academic, work/employment, and personal (see Table 2).  For example, a counselor could direct a student to a tutor, who would give reports on the student’s progress to the counselor.  Also, a counselor may have to help a student apply for a student loan so that the student doesn’t have to put in 30-40 hours on a job to support his/her education. 


2.      The data in Table 3 also lead us to conclude that counseling must be put into place for students who, for one reason or another, do poorly in their major and are told by their department or college that they can no longer enroll in this major.  These data indicate that most of the students dropping out do so in their sophomore year, the year when most students start taking courses in their major.  Now, when a student is told he/she can no longer enroll in a certain major, there is currently no adviser who will give that student help in finding another major.  That student must shop around in different departments and come up with a new major on his/her own. 


3.      The Gateway to College Learning course should be required the first semester at OU for all students with below a 3.00 high school GPA.  Students who don’t have at least a 3.00 GPA in high school have demonstrated that their study skills and/or motivation are less than they should be to succeed at the college level.  The Gateway course should also be required the second semester at OU for students who perform poorly their first semester here (2.00 GPA or lower).   


4.      The university should start collecting data on the academic intensity and quality of entering students’ high school curricula, using the U.S. Department of Education system.  This data could be very useful in the future in deciding which students have the academic preparation to succeed at OU.  The data would also be helpful in identifying students who are the most likely to have academic difficulty at OU.


One element of academic intensity is math preparation, which was observed to strongly influence the college graduation rate in a national study (Table 5).  Because of the complexity of the issue of math remediation, the task force did not have time to fully examine this topic beyond the findings given in Table 3.  Universities in several states are now grappling with new solutions to the problem of remediation, and it may be appropriate for OU to examine this issue again.  (For example, freshmen in the California State University system now must complete required remediation in English and math in their freshman year, or they are not allowed to return as sophomores; and in Florida, most remedial instruction is assigned to Florida’s community colleges.) 


cc:  President’s Graduation Rates Task Force, Nicholas Hathaway, Chair

(Nancy Mergler, Paul Bell, Doug Gaffin, Clarke Stroud, Matt Hamilton, Jeff Bloomgarden, Cheryl Jorgenson, Karen Renfroe, Gerald Gurney, and Roger Harrison)

Norman Campus Deans




Table 1 -- Retention Rates of First-Time Students



Table 2 -- Complete Withdrawal Survey



Table 3

Comparison of the Academic Records of Two Groups of 100 Randomly Selected Students Who Started at OU in 1995: 

Those Who Did Not Graduate and Those Who Did Graduate within Six Years



Students Who Did

Not Graduate

Students Who


OU cumulative GPA (avg.)




Combined cumulative GPA (avg.)




OU retention GPA after 1 year (avg.)




OU retention GPA after 2 years (avg.)




Composite ACT (avg.)




Math ACT (avg.)




English ACT (avg.)




% who took remedial math (Math 0113, 0115, or 0123)




% who made F or W in a General Education math  course (Math 1473, 1503, 1523, or 1643)




% who completed remedial math of those who took remedial math




% on academic probation (any semester)




% female




% who dropped out after:

    < 1 year

    1 year

    2 years

    3 years

    4 years

    5 years

    6 years

    7 years (may be continuing)













1.  1190 students did not graduate, 1384 students did graduate, and 137 students are continuing (total of 2711 students).

2.  Retention rates for the 1995 cohort of students based on the random sample of 100 students who did not graduate:

    One-year retention rate:  82.0%

    Two-year retention rate:  67.5% 



Table 4


High school units at the highest level of their "academic intensity"
(Adelman, 1999; 1 unit = 1 year)


Units of English


Units of math


Core lab science units


Social sci./history units


Foreign language units



Adelman, C., Answers in the toolbox:  academic intensity, attendance patterns, and bachelor's degree attainment, U. S. Department of Education, 1999.



Table 5


The Math Ladder (Adelman, 1999)


Highest math studied in high school

% of college students earning bachelor's degree







Algebra II




Algebra I



Adelman, C., Answers in the toolbox:  academic intensity, attendance patterns, and bachelor's degree attainment, U. S. Department of Education, 1999.



Table 6 -- First Semester GPA



Figures 1 and 2 -- Gateway Program

(Netscape Users: