Right?! Compare the IPEDS data collection process to the, say, US News collection process and one will quickly be disabused of the notion that government-affiliated systems are necessarily worse that wholly private ones.
I really hope this doesn’t affect the quality of the data collection process. ACTS notwithstanding, I’ve had almost exclusively positive experiences over the last 20 years with the IPEDS Data Collection System and the Help Desk in particular.
I have NO legal expertise, but is it possible that the first “state” is modifying “colleges”, “universities”, and “institutions”? Not sure what the clarity/redundancy trade-off is for these kinds of things.
If I had to guess, I’d say that they want the most recent *complete* dataset (2019 Admissions to 2025 graduation) and the most recent *available* dataset (2025 Admissions). And then everything in between.
At this point, I’ll take foolish optimism over wisened pessimism!
When I’ve asked, I’m always told, “The first year will be 7 years of data but moving forward it’ll only be 1”. But, like, only 1 of the datasets this year was “complete” (admissions through 6-year grad). Hard to believe they won’t want us to complete the other 6 years as the data becomes available?
ENDORSE
All this to say, in addition to the questionableness surrounding the ACTS data itself, rushed and ill-considered decisions may actually be making the other IPEDS data that we’ve previously come to rely on worse.
In order to fix fatal errors, schools are being directed to re-open previously submitted, well-established surveys and alter the data to match ACTS! That’s right, remove deferred students from IPEDS Admissions even though the instructions said nothing about inclusion/exclusion of deferrals…
It’s no wonder, then, that when excluding deferred first-years for ACTS, errors abound! “You reported this number before, but now you’re reporting something else. Please explain”. The explanation, of course, is that you changed a very basic reporting rule! Now this would be simply annoying, but…
You see, the First-Time, Full-Time (FTFT) class is referenced *throughout* IPEDS. In addition to Admissions, you’ll find it in Cost II, Fall Enrollment, 12-Month Enrollment, Graduation Rates, Outcome Measures, and Financial Aid. And In *none* of these surveys, does it ever mention defers…
No, NCES decided that in the absence of a formal App/Admit record in the 2024-25 admissions cycle, there should be no Fall 2025 matriculation record either! That is: Apply - No, Admit - No, Enroll - No in 2024-25. Let me tell you: this is a consequential decision…
From an IPEDS reporting perspective these additional App/Admit records have few radiating effects: the IPEDS Admissions survey is the only place where these data appear. Did NCES agree with this (reasonable) assessment? I wouldn’t be writing this thread if they did…
Seems reasonable to record the Fall 2025 matriculation and then add an App and Admit record in 2024-25. That is: Apply - Yes, Admit - Yes, Enroll - Yes. This certainly “balances the books”, perhaps at the cost of generating additional App and Admit records…
I think most people would agree on the 2023-24 record: Apply - Yes, Admit - Yes, Enroll - No. What about 2024-25? There may not have been a standard Application and Admit process despite there having been a matriculation at the end of it…
Specifically, some of the few remaining folks at NCES made ill-advised decisions regarding how to treat students who deferred enrollment. Take, for instance, a student who applied and was admitted in 2023-24 but decided to postpone matriculating until Fall 2025…
One of the things that wasn’t really mentioned in the ACTS hearing yesterday was the negative effect that ACTS has had on *already-submitted* IPEDS data…
Looks like US News is gonna have to tweak their earnings metric again…
I takes a village to produce a STEM major. At my Liberal Arts College, most majors take about 2/3rds of their courses *outside* of their major. Setting aside the benefit of a varied experience, where do folks think non-major enrollments are going to go if we keep getting rid of “low major” programs?
Honestly - and I get that this is a *me* thing - but I would settle for Dolphin being able to play Triforce MAME zip files (rather than having to get ISOs).
Sorry, did you mean “merged to *main*”? I watched the video hoping that Triforce would magically be a part of .287, but that doesn’t seem to be the case?
ACTS continues to be the greatest example of why TRPs are absolutely essential.
In case you were interested in the full provenance of the super standout feature… https://www.reddit.com/r/MAME/s/pIgCLsp2BW
I don’t think your frame is narrow. Honestly, “having AI go over it” was my charitable version of the interactions with AI that colleagues have described to me. Personally, though, I haven’t seen a compelling use case.
Maybe this is not completed related but I think I there is something different about creating something yourself first (without AI cues) and then maybe have AI go over it vs having AI create it first and then a human going over it. I think the former will lead to a superior product.
…like if I ask you to tell me all 50 states, you’ll be better at recalling a given subset of 20 if I give you no cues as opposed to giving you the other 30 states to start…
There’s a concept in cognitive psych called “part-set cuing inhibition” which demonstrates that it’s harder to remember a subset of a list when you’re cued with a different subset vs no cues at all…
There’s a larger point to be made about one person’s “inefficiency” being another person’s “thoughtfulness”, but I’ll let someone more eloquent make it!
I think you have to have an AIR account? But here it is: hub.airweb.org/groups/commu...
There may be others, but this is the one I was referring to.
A classic case of “building the plane as you fly it”. It really hammers home why new surveys often take years to be put into the rotation.