Demonstrated Interest (DI) vs Likelihood to Enroll (LTE)

[Anonymous]

For colleges who say they don’t consider demonstrated interest on their CDS, do they still consider or calculate LTE? If they don’t consider DI, does that mean they don’t use computer metrics to consider how likely you are to enroll? And how does yield fit in? If a college says it doesn’t consider demonstrated interest, but still appears to yield protect, does that mean they don’t calculate DI but might still consider LTE? Based on stats? I realize this is an inside baseball type question but thought someone here might know. Thanks!

[Anonymous]

I'm guessing yes, because DI is only one part of LTE:

"The predictive step of algorithmic enrollment is aimed at estimating how likely an accepted applicant is to enroll in a specific college. To do this, a college will first consolidate data about a college’s past applicants, including variables like their high school GPA, standardized tests scores, FAFSA data, how much financial aid they received, where they live, and demographic information. Colleges also frequently incorporate engagement metrics, such as how often applicants attend college recruitment events and what percentage of college emails they read. Using this historical admissions data, the college or a vendor will then build a predictive model with these variables to predict whether each of the accepted applicants chose to enroll."

https://www.brookings.edu/research/enrollment-algorithms-are-contributing-to-the-crises-of-higher-education/

[Anonymous]

This is the kind of thing you won't know because you aren't entitled to know the ins and outs of a school's enrollment management strategy.

[Anonymous]

These programs/consulting firms are very expensive and many schools can't afford them.

[Anonymous]

Anonymous wrote:These programs/consulting firms are very expensive and many schools can't afford them.

"According to a 2015 Educause Survey, over 75% of colleges and universities use analytics for enrollment management, up from just over 60% in 2012, making it the most common form of data analytics in higher education."

https://www.brookings.edu/research/enrollment-algorithms-are-contributing-to-the-crises-of-higher-education/

[Anonymous]

Anonymous wrote:This is the kind of thing you won't know because you aren't entitled to know the ins and outs of a school's enrollment management strategy.

I would think that a citizen of a state where the public schools uses these algorithms should be entitled to know

[Anonymous]

Anonymous wrote:I'm guessing yes, because DI is only one part of LTE:

"The predictive step of algorithmic enrollment is aimed at estimating how likely an accepted applicant is to enroll in a specific college. To do this, a college will first consolidate data about a college’s past applicants, including variables like their high school GPA, standardized tests scores, FAFSA data, how much financial aid they received, where they live, and demographic information. Colleges also frequently incorporate engagement metrics, such as how often applicants attend college recruitment events and what percentage of college emails they read. Using this historical admissions data, the college or a vendor will then build a predictive model with these variables to predict whether each of the accepted applicants chose to enroll."

https://www.brookings.edu/research/enrollment-algorithms-are-contributing-to-the-crises-of-higher-education/

Thanks! This makes sense to me. An applicant has some control over their DI. Does any applicant have any means to impact the LTE algorithms (aside from DI)?

[Anonymous]

Anonymous wrote:

Anonymous wrote:These programs/consulting firms are very expensive and many schools can't afford them.

"According to a 2015 Educause Survey, over 75% of colleges and universities use analytics for enrollment management, up from just over 60% in 2012, making it the most common form of data analytics in higher education."

https://www.brookings.edu/research/enrollment-algorithms-are-contributing-to-the-crises-of-higher-education/

+1 I am a professor at an R1 public. Last year, I attended an admissions committee meeting with a vendor, Slate, which seems to be one of the dominant companies in the industry. I can't provide a link but the current percentage of US colleges and universities that use analytics for enrollment management is now 85%. However, many schools still use the essential features to manage applications. I noted that from the list of R1's and top 100 schools, all use enrollment management software AND data mining algorithms to determine yield and LTE.

[Anonymous]

Interesting! What kind of data is being mined?

[Anonymous]

Anonymous wrote:Interesting! What kind of data is being mined?

PP/Professor here: If you have time (it's 50 mins), watch this video from a vendor: https://youtu.be/gHALuwbmJFw

You can skip to the presentation of data from 3 schools.

[Anonymous]

My theory: test optional threw such a wrench into yield algorithms for college class of 2025 not only because the algorithms used scores, but they didn't seem to realize that test optional applicants were more likely to yield and they misjudged the admission opportunities of high-score submitters at more selective schools in the new test--optional scenario. Essentially, their mathematical modeling based on the past was less relevant to predicting the future.

I wonder if they improved (more accurately predicted yield) with class of 2026. Anyone know if many top-100s are overenrolled again?

[Anonymous]

Well, I suppose that explains a lot.

[Anonymous]

Anonymous wrote:

Anonymous wrote:These programs/consulting firms are very expensive and many schools can't afford them.

"According to a 2015 Educause Survey, over 75% of colleges and universities use analytics for enrollment management, up from just over 60% in 2012, making it the most common form of data analytics in higher education."

https://www.brookings.edu/research/enrollment-algorithms-are-contributing-to-the-crises-of-higher-education/

Using "analytics" is a pretty generic term. That's not the same thing as hiring a firm that does predictive modeling.

A predictive model is a very specific thing and it's costly.

I know from experience.

[Anonymous]

Someone posted a you tube link of essentially a zoom meeting pitching their services - probably was removed above. Anyway, it's quite interesting - individual likelihood of attending score. Still listening, but wondering whether, and how, this score would fit into an admission decision.

[Anonymous]

Someone needs to reverse-engineer the algorithms and post the findings to help with transparency and level the playing field.