"College applicants should understand how schools use predictive analytics and what it means for admissions decisions.
DATA IS EVERYWHERE IN A connected society. Data may determine what programs a streaming service queues up, the tailored ads that appear on websites, or personalized health and fitness plans. Data may also play a factor in whether an applicant is accepted into his or her top college choice. In short, data is everywhere, including in algorithms for college admissions.
‘Data drives everything,’ Angel B. Pérez, CEO of the National Association for College Admission Counseling, wrote in an email. ‘Enrollment officers must report average GPA, average test scores, revenue, discount rates, financial aid dollars awarded, ethnic breakdown…the list is endless. Enrollment officers are steeped in data.’" …
“After academic achievements and the essay or writing sample, colleges considered demonstrated interest as the next most important factor in making an admissions decision, even over teacher and counselor recommendations, extracurricular activities and class rank.”
The source for that statement is NACAC’s 2019 State of College Admission report (linked in the article).
Table 7 on Page 15 contains the ranking of admission factors based on an N of 220ish…and it does have demonstrated interest ranked ahead of those other factors.
Do you have any reason to doubt the veracity of that data? I think there are about 1,200 NACAC member colleges, so not a huge sample, but for this particular sample that’s what the data show.
There are plenty of colleges that say that demonstrated interest is considered. There have actually been studies that show the correlation between visits and improved yield.
I certainly agree with @RichInPitt’s sentiment expressed above in post #1.
The chart on page 15 of 28 pages of the 2019 NECAC report does support the article’s statement, but the percentage of participating schools which rank “demonstrated interest” as a most important factor in admissions is just slightly over 16%.
Not sure about the confusion regarding “incompetence”, but it refers to a state of being without ability, knowledge, fitness, etc. regarding a particular matter.
Regarding demonstrated interest, the specific numbers were as listed below this post. 1/3 of surveyed colleges said demonstrated interest had no significant influence on admission decisions, and the clear majority said it had either no influence or limited influence. In contrast. the clear majority of colleges said counselor LORs had at least moderate importance in admission decisions, and only 18% said counselor LORs had no influence. Teacher LORs were similar.
I think most would interpret this as colleges as a whole treat LORs as more influential than demonstrated interest (conflicting with article quote). I think the more important distinction is that there are a huge number of exceptions, with many different admission systems at different colleges, rather than a single weighting system. Some colleges really emphasize demonstrated interest in admission decisions. Some colleges do not consider demonstrated interest at all. Looking at the regression results in the appendix, lower acceptance rate colleges were more likely to say they emphasized LORs, but were slightly less likely to say they emphasized demonstrated interest. The focus of this forum tends to be lower acceptance rate colleges, so among colleges frequently discussed on this forum, one would expect a different pattern than the overall averages.
The issue with the CDS is that institutional reporting fills this out, not admissions, and some schools don’t have good coordination between those departments.
It is not uncommon that sections of the CDS might be out of date, or wrong (e.g. section c7, importance of admission factors), or different than how admissions does it (e.g. GPA calc)
Yes, any one college’s CDS could have errors. But how likely is it that hundreds or thousands of colleges have CDS errors in the same direction that understate the importance of level of interest and/or overstate the importance of recommendations, essay, class rank, extracurriculars, etc.?
One thing that is very important at most colleges (and another data point) but not listed above is the ability of the student to be full pay versus one than needs financial aid. For most colleges, this is going to be high on the list of data of what is important in the application, unfortunately.
Most colleges are need blind for individual applicants in terms of whether admissions considers the financial aid information of the individual applicants, even though they may choose to emphasize applicant characteristics that tip the overall class toward a desired tuition yield. For example, favoring legacies tends to tip the class toward higher tuition yield, even though a few individual legacies may have high financial aid need.
Ability to pay was one of the questions they asked about in the NACAC survey. The surveyed colleges reported the following, making ability to pay the least influential of the 20 surveyed criteria. It’s certainly possible that the colleges were not being truthful, although the names of the colleges are not published, reducing the risk of negative repercussions for unpopular responses.
Ability to Pay
No Influence – 81%
Limited Influence – 14%
Moderate Influence – 4%
Considerable Influence – 1%
If ordered from most to least average influence, the survey order would be similar to the following:
Most to Least Reported Average Influence: NACAC Survey
Grades
Strength of Curriculum
SAT/ACT Scores
Essays
LORs
ECs
Demonstrated Interest
Class Rank
Work
First Gen
AP/IB Test Scores
High School Attended
State of Residence
Interview
Race/Ethnicity
Legacy
Gender
State Graduation Exam Scores
SAT II Scores
Ability to Pay
A comparison to UCB’s CDS survey. They both agreed that HS GPA, HS course rigor and standardized tests were the 3 most influential factors by far, in that order. They also both agreed that the next 2 most influential were essays and LORs. Among the top 8, the only one with a large difference appears to be demonstrated interest. NACAC had it above rank and well above interview. CDS had it below both rank and interview. There are many possible reasons for this difference, including things like the wording of the question and whether it includes applying ED, and different sampled colleges.
Most to Least Reported Average Influence: CDS Survey
Colleges are big business. Are we really to believe that colleges don’t really care about that. Hog wash. Maybe the intitial adcom reviews don’t consider ability to pay but certainly when they build a class they absolutely take that into consideration. If colleges are using algorithms for other data, you can be rest assured that they will have x number of full pay students, year in and year out otherwise these colleges would be losing money every year.
Being need blind or placing little emphasis on ability to pay doesn’t mean the college must meet the full need of the lower/middle income students that they do admit. Most colleges don’t. At many colleges, the majority of students take out loans to cover costs.
Many of the elite private colleges are smart enough to be able to build a class that targets at specific level of FA need, while being need-blind for individual applicants.
It is not hard to see that a college can tip the class to less FA need by doing the following, without having to look whether individual applicants applied for FA or how much FA they are likely to need.
Filling more of the class in early rounds.
Increasing weight toward standardized test scores relative to HS GPA.
Increasing weight toward legacy.
Increasing weight toward athletic performance, particularly in "preppy sports".
Favoring high SES high schools in recruiting and admissions.
Requiring non-custodial parent information for FA (probably eliminates about half of potential FA-needing students).
Of course, far more numerous are the colleges that just do not give good FA. They may admit FA-needy applicants on a need-blind basis, but know that their yield will be low because they will not find those colleges to be affordable.
"It’s certainly possible that the colleges were not being truthful’
I do think there’s amount of misleading going on, even if it’s an anonymous survey. However It’s more likely that adcoms think they’re being need blind when they’re not. The factors that UCB noted that favor high-SES students is probably an implicit bias, so adcoms accept kids in wealthy zip codes think wealth of the family had nothing to do with it.
However, even Harvard has a budget and most if not all colleges would like a certain amount of the class to fund others, and then come in with some FA. I’m going to guess the best have of a certain percentage of the class subsidize the others.