I developed a small app to calculate T14 admissions chances.
https://schoolpredict.com
A friend had asked about what could make their profile strongest for a specific school, which led to a grad project last year doing analysis for them. I used a portion of the project, made it more general, and hosted it on the site.
It takes more factors into consideration than just LSAT, GPA, URM, and ED -- such as Non-Traditional, years out, and so on. I only did the T14 because of the amount of effort it took to build the individual models and clean the data, but I will expand to more schools in the near future.
I would welcome any feedback or suggestions for new features.
T14 Law School Admissions Calculator Forum
-
- Posts: 97
- Joined: Mon Jul 24, 2017 11:30 pm
Re: T14 Law School Admissions Calculator
Interesting tool! Unfortunately, as an international with a 3.71 uGPA (top 5% from my school and translated as a "superior" by LSAC) I don't think it is useful in my case because my 3.71 GPA would be considered lower in the U.S. scale, and the fact that the GPA appears to be a letter grade is perhaps not factored in the calculation even though I checked the box of international..
-
- Posts: 3594
- Joined: Mon Aug 07, 2017 9:52 am
Re: T14 Law School Admissions Calculator
Did you find that - Yale and Stanford aside - factors beyond LSAT/GPA/URM status actually have a substantial impact on one's chances of admission? How do you define "non-traditional"? I ask because all of our experience, as well as MyLSN's data, suggest that it's almost purely a numbers game. Splitters' cycles are less predictable (as one would expect), but non-splitters' admissions and scholarship offers almost always line up perfectly with what MyLSN predicts based on nothing more than LSAT/GPA/URM status.jgroland wrote:It takes more factors into consideration than just LSAT, GPA, URM, and ED -- such as Non-Traditional, years out, and so on. I only did the T14 because of the amount of effort it took to build the individual models and clean the data, but I will expand to more schools in the near future.
-
- Posts: 2
- Joined: Thu Mar 07, 2019 5:57 pm
Re: T14 Law School Admissions Calculator
Yes, this is a limitation from the Law School Numbers data. Since all that was available was the reported GPA, it wasn't possible to translate each record's GPA into a class rank. When I was building the models I did scale everything to be on the 4.0 scale. All of the factors like "international student" or "military" are factored as binary variables but don't directly affect the GPA scaling.Lxwind wrote:Interesting tool! Unfortunately, as an international with a 3.71 uGPA (top 5% from my school and translated as a "superior" by LSAC) I don't think it is useful in my case because my 3.71 GPA would be considered lower in the U.S. scale, and the fact that the GPA appears to be a letter grade is perhaps not factored in the calculation even though I checked the box of international..
GPA and LSAT were definitely dominating factors. URM status and ED were the next most important factors, which probably is in line with everybody's experience. Having years out of school did have a big impact in some cases, and my assumption is that this is a proxy for "related work experience." There is no way to distinguish between years out of school and years of legal work experience with the data. It's possible that having years of useful work experience improves admissions chances for a candidate with weaker scores.QContinuum wrote:Did you find that - Yale and Stanford aside - factors beyond LSAT/GPA/URM status actually have a substantial impact on one's chances of admission? How do you define "non-traditional"? I ask because all of our experience, as well as MyLSN's data, suggest that it's almost purely a numbers game. Splitters' cycles are less predictable (as one would expect), but non-splitters' admissions and scholarship offers almost always line up perfectly with what MyLSN predicts based on nothing more than LSAT/GPA/URM status.jgroland wrote:It takes more factors into consideration than just LSAT, GPA, URM, and ED -- such as Non-Traditional, years out, and so on. I only did the T14 because of the amount of effort it took to build the individual models and clean the data, but I will expand to more schools in the near future.
The factors that led to higher admissions probability for some schools could lead to lower admissions probability for other schools. I had assumed it would be best to just maximize all stats across the board, but it wasn't always the case. Cornell didn't seem to like TFA experience, but this could be because the candidates with TFA experience applying to Cornell happened to have worse LSAT/GPAs. The specific self-identified race did have an impact. Self-identifying as LGBT+ generally seemed to be beneficial. Having a fee waiver did negatively impact admissions probability in some cases. It's hard for me to make generalizations because each school appeared to group candidates into different admissible categories/clusters that are unique from school to school.
I included the field "Non-Traditional" if it was marked as such on the LSN profile. I'm not sure what information a self-identified "Non-Traditional" status can convey, but I think it could convey several things. E.g., the person was out of school for some years, worked in an unrelated field, started undergrad later, or took several additional semesters to finish. I'd be speculating to pin down an exact definition, other than it being a flag for having an atypical profile. There is some overlap with the "years out of school" factor. Because the data is non-specific -- it's hard to say exactly what the status conveys.
Want to continue reading?
Register now to search topics and post comments!
Absolutely FREE!
Already a member? Login