Hmm, what do you have in mind, because I have seen the opposite. For example:
Q: So, you know, I have to ask, what, what data are we talking about here? Is this GPA testing?
A: So we, yeah, so we’re using academics. So we’re using 64 different combinations of academic data, depending on what’s available in a school. And we weight it accordingly. So, you know, things like the obvious things like is or gpa, is it weighted or not? Is is there a way of assessing the rigor of curriculum in the high school? Is there class rank? Is there testing? What kind? And the combinations of those elements around academic achievement give us a way of taking a big pool and starting to sort it and say, okay, let’s start at the top and work our way through it.
And as opposed to reading them one by one without any intentionality. And, you know, you may be reading the wrong ones first. And so we’ve, we’ve kind of, you asked a question about, you know, volume and how we’ve had to rethink the way we do our work. This is an example of that where we’ve had to pivot a little bit from the old fashioned way of starting and just documenting and reading everything to say, okay, let’s, let’s be clear at the beginning about where to begin.
Q: Do you take any institutional history into consideration? Like some schools, when they have a track record of a lot of students from a school, they’ve, you know. . . . In first year class from this school, any of that kind of stuff?
A: No. So we’re, we’re not doing anything historical. We are context comes into it. So we are adding contextual elements from the school. So when there’s profile data that we can plug into one of the logic trees to say X percent go to a four year college, here’s the mean SAT against the student submitted score. We’re taking pre-populated information and we’re putting it through an algorithm that helps us sort the pool upfront. And it saved us, this year was the first year we did it, it saved us about three weeks of work. . . . And let us get into the reading process much sooner than we were able to do in other years when we were manually going one by one through all those files with admission people doing it. And also it just made it less open to misinterpretation by various admission officers. We were, you know, the, the, an admission officer could override the algorithm, but that was rare.
Q: Did did you come up with the 64 criteria by like brainstorming as an admission office? . . . Was this the work of consultants that, you know, you outsourced? How did you contact with that?
A: No, we, no, we insourced it. This was our senior admission officers last summer, had a retreat and I had a working group that kind of plotted through it and said like, what is available to us on the secondary school report on a transcript and testing through the college board landscape, like what’s out there that we could glean from the documents submitted, populate our database, and then the logic tree kicks in and says, okay, you know, these elements are present. This is algorithm 27, whoop. And it came up with an academic assessment for us.
Q: As you network with other peers at other schools. How common is this Lee? How many, you know . . .
A: I think I’m a pioneer.
Q: Yeah. 'cause I am not hearing this. Yeah. So I’m like, I’m much out of the loop on this. I mean, I’m . . .
A: No, no, I, I was sharing this with some fellow deans recently and they were like, wow. And I, I just, you know, it was a giggle because I’m, I was a humanities student, but I, you know, the data is there for us to use if we can. And what, in this example I’m sharing is me trying to think creatively about what do I know, how can I have the technology we have help tee up the work, not do the work. This isn’t a chat GPT version of reading a file, but help prep it so that the admission officer can go in and more elegantly use the information that’s been submitted and spend time where we need to spend time.
Q: Well, anything that saves three weeks. Yeah. I have a feeling you’re gonna have some other schools reaching out to you and asking for some Tips.
A: Yeah, yeah, yeah.
Q: This is where you tell Dartmouth, Hey let, let me get some consulting money on the side.
A: Yeah. Really. Let’s do a, let’s do a patent. No, but it was, you know, all of it is, you know, none of it was invented. It’s all there in the secondary school report. It’s, you know, being able to just load it into your computer and have the power of technology help map it for us.
Q: I do think it’s important for us to let our listers know though, that you are not just making decisions off of this. You’re still doing a full holistic read with the human element. Bringing in aspects of nuance that cannot be captured by landscape or, you know, secondary.
A: No, we don’t do this school profiles. No, no, no. What I’m describing doesn’t touch an essay. Yeah. Or a recommendation or an interview. It’s, you know, it’s, it’s taking what was the first step in our reading process, which is the academic review and automating it to a degree that lets the admission officer then go in and say, okay, this has been calculated for me. Does it seem accurate? If not, why? Or, oh, there’s a new piece of information, I’m gonna plug it in. And the, it’s adaptive. So the score adjusted as we added new, you know, when we see Julia’s letter saying if we ranked, she would be number three. You can plug that in and that changes the algorithm.
