The NCAA: If You Want to Compete, Stay in School

The dataset we are working with is about academic scores for NCAA athletic programs. There are 6,511 observations corresponding to an NCAA sports team at a particular college. The dataset contains variables such as school name, sport name, conference, school type (private or public), an academic score for each year (from 2004 to 2014), a four-year retention rate for each year (from 2004 to 2014), and the proportion of athletes that are elgible for each year (from 2004 to 2014). The academic score is called the Academic Progress Rate (APR). It is intended as a metric for the NCAA to hold academic institutions accountable for the academic progress of their student-athletes. The APR is calculated by giving each student-athlete 1 point for staying in school and 1 point for maintaining academic eligibility, then dividing by the total number of possible points and multiplying by 1000. If this number (as a 4-year average) drops below 930, then the team is subject to penalties of varying degrees such as practice and competition reductions, financial aid reductions, and coaching suspensions. This rule was first implemented in 2004.

Our main focus of the project is to determine what factors are associated with APR scores. In order to achieve this, we came up with several research questions. These include: what sports or schools have lower APR scores on average? Is there a relationship between the number of atheletes on a team and APR scores? Do APR scores vary by location? Is there a trend in APR scores over time across different sports and schools?

As previously mentioned, when teams earn an APR score less than 930, they could be subject to penalties by the NCAA if they fail to bring the score up over the course of a few years. Therefore, we thought it would be interesting to examine the sports that were most frequently scored under this threshold in 2014. This would help us answer the question of which sports are most in danger of being penalized for low APR scores (as of 2014). We calculated the proportions of each sport that earn scores less than 930.

The graph has been limited to show the 9 sports with the highest proportions because there are 38 sports in the dataset, and we thought it would be more interesting (and less cluttered) to look at a smaller subset of sports. The visualization suggests that the pictured sports have the highest proportions of teams with APR scores lower than 930 in 2014. Thus, they are the sports that are most frequently at risk of being subject to penalties by the NCAA for the academic progress of student-athletes if they continue to score low. For example, 12% of men’s basketball teams across all schools fall below the 930 threshold. To be more specific, 42 out of 350 men’s basketball teams earned a score less than 930 in 2014. It is interesting that all different sorts of sports occcur in the subset. For example, it does not appear to be only team or individual sports. It is also interesting that there are more men’s sports with proportions greater than 8% than women’s. We will focus on these sports for the remainder of our analysis.

We also thought it would be interesting to examine the schools that have the most teams with scores less than 930. This would help to answer the question of which schools are at risk for being penalized for having low APR scores. We calculated the proportions of schools with teams that earned scores less than 930 in 2014.

The graph has been limited to show the 11 schools with the most teams with APR scores lower than 930 in 2014 since there are 385 schools in the dataset. The barplot suggests that these are the schools with the most sports teams at risk for being penalized by the NCAA for the academic progress of their student-athletes in 2014. For example, 73.3% of sports teams at Savannah State University had APR scores less than 930 in 2014. More specifically, 11 out of 15 athletic teams at Savannah State University had APR scores less than 930. Interestingly, many of the schools are in the South. It is also interesting that there are 11 schools in the dataset that had at least one third of their teams falling below the 930 threshold in 2014. These are the schools we will focus on for the remainder of our analysis.

In order to deduce which factors or variables are significant to predicting a sport team’s academic score, we examined the relationships between the APR score of 2014 and the different sports, the total number of athletes on the sport team, and the type of college/university.

The above graph suggests the sport teams that have the relatively lower scores are mostly from public universities, which is most evident in Men’s Basketball and Men’s Cross Country. It is also seems that the sport plays a role in academic score, because the graph demonstrates that the Men’s Basketball, Men’s Cross Country, Men’s Track, Indoor, and Men’s Track Outdoor have the largest range of academic scores and have a lot more individual teams with relatively lower scores. This graph also helps us deduce that the number of athletes on a sports team likely does not have a significant association with the academic score.

Motivated by the bar plot, which seemed to suggest that many of the schools that struggle with low APR scores are located in the South, we wanted to explore to what extent geographic location is associated with score. Using schools’ Latidude and Longitude along with the average 2014 academic score of all the teams within each school, we created the following plot:

The mapped data overviews the distrubtion of academic scores across the country at the school level. This map highlights that a majority of the schools in the dataset are located in the east and eastern mid-west. It appears that there is not a large variance in average score among schools, but there are some interesting trends. Most schools generally have high acamedmic scores among their teams, and there are certain clusters in the US where average scores seem particularly high. While schools in the Northeast have average scores that are close to 1000, the schools in the South appear to generally have lower scores, on average.

Motivated by the sports that have a particularly high proportion of teams that don’t hit the 930 APR-score cutoff for preventing restrictions on funding, we decided to explore how that APR score has changed over time, on average, for those sports. In particular, the following time-series graph plots the average APR score for each of the given low-score sports, for each year from 2004 to 2014.

Because the NCAA enforcing the 930 APR-score rule in 2004, it makes sense that many of these sports showed a general increase in APR score over time. However, some of these sports appear to have improved more than others. In particular, Men’s Basketball, Men’s Cross Country, Men’s Tennis, Men’s Indoor and Outdoor track, and Men’s Wrestling all appear to show steady improvements in APR score over time. However, Mixed Rifle and Women’s Bowling appear to fluctuate more in their improvements in average APR score over time, with perhaps a less predictable upward trend. While Women’s Fencing does not appear to have improved their APR score over time, it is already quite high and does not have much room for improvement.

Because we noticed that some schools tend to have a particularly high proportion of their sports teams’ APR scores below 930, we also decided to explore time-series relationships with those schools:

It appears that Alabama State’s teams had APR scores of about 900, until the year 2009, when the scores started to gradually improve. On the other hand, most of these other schools, such as Tennessee State University and Prairie View A&M University, appear to have maintained similar APR scores over time. A third category of schools, including Florida A&M University, Mississippi Valley State University, and University of New Orleans, shows much more fluctuation in the APR score from year to year. Years with higher or lower APR scores could correlate to motions put in place by that university’s Athletics Department (or lack thereof), but further investigation is required to know for sure. University of South Carolina Upstate appears to have data only as early as 2008, and Grand Canyon

University and Abilene Christian University appear to have data only as early as 2014, probably because these are the respective first years each university had a sports program. However, it is interesting that some of the schools that are struggling with having high APR scores are schools with newer sports programs.

Of the schools and sports for which we have considered APR score versus time, it appears that many sports have a similar gradually increasing trend over time. However, from school to school, there appears to be more variation in this trend. Therefore, to get a sense of the trajectory of academic success (as measured by APR score) of a particular NCAA sports team, it may be sensible to consider the relationship of APR over time faceted by school before the same relationship faceted by sport.

In summary, given this dataset of NCAA sports teams, we investigated various relationships between APR score, a metric the NCAA uses as an indicator of academic performance, and other variables of a team, such as sport, team size, school type, location, and year. One result we found was that some sports tend to have lower APR scores than others on average; however, most sports’ average APR scores have improved steadily between the years of 2004 and 2014. On the other hand, we found that many of the schools with low APR scores on average are located in the South, and how their APR scores have changed between the years of 2004 and 2014 depends heavily on the school.

Since the NCAA 930 APR score rule was implemented in 2004, many sports have shown an increase in their teams’ APR scores. This fact illustrates an academic improvement among sports teams in the NCAA. Or does it? It certainly illustrates an academic improvement if the definition of academic achievement is solely APR score, but there are many other possible definitions of academic achievement.

One advantage of using the APR score as a metric is that it is very simple. But is it too simple? If the NCAA used a different metric, such as GPA or post-graduation salary, would we be able to learn something more about how to foster athletic programs that nourish people as both students and athletes? What about the fields of study the student-athletes choose to pursue? Further research could explore these questions.