X

Week 4: March Madness & Mobility

Apr 04, 2021

Welcome back to my blog everyone! With March Madness going on, I thought about ways I could connect the tournament to my senior project. The tournament brings both excitement and disappointment for fans and for people trying to get the perfect bracket. I would’ve had a perfect bracket if Oral Roberts didn’t beat Ohio State in the Round of 64…but that’s beside the point. March Madness is also where a team can write their Cinderella story and become a lower seed to make the Final Four. This year, that team was UCLA. Unfortunately for them, they lost on an unbelievable shot by Gonzaga at the end of overtime last night, but they proved that anything can happen in March!

I started this mini project a while ago to see if a college’s basketball standing in March Madness would affect the number of applications to that college. Listed below are my hypothesis, procedure, statistical analysis, and conclusion. It is pretty long, so you can skip to the Discussion if you want.

Hypothesis:

My null hypothesis is as follows: If a college’s March Madness standing increases, then their access rates, and therefore social mobility, will increase. I will be tracking Duke University as it is one of the few private universities (which seem to have lower access rates) that have been selected for the March Madness tournament nearly every year. I will also be tracking Northwestern and UVA, two universities that have not shared the same March Madness success but had solid athletics. I was hoping to establish a causation between the two variables (March Madness standing and access rates). I would need the access rates of the colleges for each year between 2000 and 2002, but I only have the average access rate over that time period. I would also need the data regarding the number of students from the bottom 20% of the income distribution who applied to these colleges, so I truly see if the access rates are being affected. Due to these limitations, I will focus on making a correlation.

Procedure:

I selected Duke University and noted their March Madness standings between the years 2000 (Starting in 2000 since data is not available for 1999) to 2002 as they were in the tournament each year. I selected Northwestern University, also a top-tier D1 school but was not selected to participate in March Madness between these years and proceeded with the same steps. Lastly, I selected University of Virginia, a team with relative success, and proceeded with the same steps. I compared the number of applications to the March Madness Performance for all the schools to look for any trends.

Statistical Analysis:

Duke:

2000 (Class of 2005):

Duke Applications: 14,468

March Madness Performance: National Champions (2001)

2001 (Class of 2006):

Duke Applications: 14,647

March Madness Performance: Sweet Sixteen (2002)

2002 (Class of 2007):

Duke Applications: 15,860

March Madness Performance: Sweet Sixteen (2003)

Northwestern:

1999 (Class of 2004):

Northwestern Applications: 15,460

March Madness Performance: Not in March Madness

2000 (Class of 2005):

Northwestern Applications: 14,725

March Madness Performance: Not in March Madness

2001 (Class of 2006):

Northwestern Applications: 13,987

March Madness Performance: Not in March Madness

UVA:

1999 (Class of 2004):

UVA Applications: 19,670

March Madness Performance: First Round

2000 (Class of 2005):

UVA Applications: 16,656

March Madness Performance: Not in March Madness (NIT First Round)

2001 (Class of 2006):

UVA Applications: 15,052

March Madness Performance: Not in March Madness (NIT Second Round)

Discussion:

One important note to make here is that the March Madness tournament started after all applications to the university were submitted. The NIT is the tournament for teams not selected to participate in March Madness. We can see a slight correlation between the number of applications and March Madness Performance for Duke. Duke made the tournament all three years and saw an increase in applications each year. Perhaps after Duke won the national championship in the 2000-2001 season, more people wanted to apply. It is interesting to note the sudden increase in applications from 2001 to 2002, despite Duke finishing worse than the year prior. In the blog regarding the 2003 admissions, the dean of admissions says that one reason for an increase in applications (15,860 to 16,656) is due to the need-based financial aid offered to international students.

Northwestern serves as a control group in this case. They made the March Madness tournament in 1999 (months before the application deadline), but they didn’t make another appearance until 2009. They saw a sharp decline in their applications, going from 15,460 to 13,987 in just two years.

UVA is interesting because they were decently competitive, either participating in the March Madness or NIT Tournament each year from 1999 to 2001. However, the number of applications were sharply decreasing despite the relative success of the basketball team.

Looking at the enrollment at Duke, for instance, for undergraduates during this time (since he tracks students when ages 19-22), there were 6,202 students in 2000, 5,990 students in 2001, and 5,916 students in 2002. Duke’s access rate according to Raj Chetty’s data is 3.9%, meaning that 241 students in 2000 were from the bottom 20% of the income distribution, 233 students in 2001, and 230 students in 2002. For comparison today, the 2019 enrollment was 6,526, which means 254 students were from the bottom 20% if the access rates stayed relatively constant.

Thanks to anybody who read to the end! I didn’t really expect you to, so bonus points for you I guess. See y’all next week!

References:

https://today.duke.edu/2001/04/admits406.html#:~:text=The%2014%2C647%20applications%20in%202001,of%20steady%20growth%2C%20Guttentag%20said.

https://today.duke.edu/2002/04/admits0402.html

https://today.duke.edu/2003/02/apply0218.html

https://en.wikipedia.org/wiki/Duke_Blue_Devils_men%27s_basketball#Results_by_season_(1980%E2%80%93present)

https://library.duke.edu/rubenstein/uarchives/history/articles/statistics

https://opportunityinsights.org/wp-content/uploads/2018/03/coll_mrc_paper.pdf

https://repository.upenn.edu/cgi/viewcontent.cgi?article=1371&context=fnce_papers

https://www.weforum.org/reports/global-social-mobility-index-2020-why-economies-benefit-from-fixing-inequality

https://www.piie.com/blogs/realtime-economic-issues-watch/denmark-new-american-dream#:~:text=According%20to%20the%20OECD%2C%20high,countries%20where%20inequality%20is%20low.

https://www.adminplan.northwestern.edu/ir/data-book/v51/2.01-undergraduate-admissions-statistics.pdf

https://en.wikipedia.org/wiki/Northwestern_Wildcats_men%27s_basketball#NCAA_Division_I_tournament_results

https://www2.virginia.edu/presidentemeritus/report01/growth.html

https://www2.virginia.edu/presidentemeritus/report01/uva2001-02.html

https://en.wikipedia.org/wiki/Virginia_Cavaliers_men%27s_basketball#2018:_Calm_before_the_storm

 

3 Replies to “Week 4: March Madness & Mobility”

  1. Eric M. says:

    Interesting! If the applications were submitted before March Madness even starts, do you think that any reasonable causation can be determined in the first place?

    1. Neel D. says:

      Well all the applications are submitted before March Madness starts (assuming the standard November and January deadlines), but it’s more of the fact that someone applied because of the college’s success last year and their predicted success this year. However this can vary, which is why I don’t think a causation can be made. For example, Duke missed the tournament after years of being in it, but I don’t think applications will now decrease because of one bad year from the basketball team.

  2. Mr. Loomis says:

    Keep in mind it is very difficult to establish any type of causation via an observational study. A well designed experiment in which some schools would be randomly assigned to do well at the NCAA tournament and some schools would be randomly assigned to do poorly at the tournament (or not appear at all) would be needed in order to establish any type of causal relationship between success at the tournament and number of applications.

Leave a Reply