Thread by @akilbello

My latest article explores the history of and why it's time to end the of the 51 year SAT/ACT "experiment" at the University of California. lnkd.in/grvz7apt

A few interesting tidbits I didn't add to the article.

1950s - the University was confronting serious problems associated with growth. The GI Bill had increased enrollments significantly and the baby boom generation had entered the school system. The goal was 12.5% eligibility

1960 - BOARS concluded that the study did not indicate any additional predictive power associated with the SAT. A subsequent study of showed a better correlation with FGPA, but not enough to convince the Academic Assembly of the value of adopting an admissions test requirement

1965 - studies estimated that the University was significantly out of compliance w/ the eligibility cap of 12.5% called for by the CA Master Plan for Ed. Adopting an admissions test requirement was seen as a “relatively easy” means of reducing the size of the pool.

Seriously, the UCs have used the SAT and ACT as simple tools to manage enrollment. They might as well have filtered by head size.

It was a mildly academic mildly preldictive mildly valid politicaly justifiable measure that let them reduce eligible applicants.

Let's not forget that during the 60s @CollegeBoard establish its first west coast office in order to lobby the UCs to adopt the test. they called it #OperationGoldenBear

https://twitter.com/akilbello/status/1215857557316739072?s=20

@CollegeBoard another interesting source of history is The Case Against the SAT

Holy #PairedText, Batman!

ACT's former CEO was recycling ETS analogies from 1984

google.com/books/edition/…

Let's also not forget there were major changes in the test. How can 5 versions of 2 different tests be identically predictive?

erikthered.com/tutor/sat-act-…

I'm fascinated by how little discussion the media (and even the Regents meeting) gave to the fact that almost every change in the Eligibility index and test use has been about limiting the pool of students not about predicting who can do the work.

1992 - more limiting

In 2001, UC president had a lot to say about the tests which forced CB to integrate the SAT II: Writing into the SAT I and to do major changes to the SAT I.

https://twitter.com/akilbello/status/1224879053083226114?s=20

If you haven't read Atkinson's words you should

rca.ucsd.edu/speeches/Colle…

rca.ucsd.edu/speeches/aera_…

the ACT research I reference is the discrepant data that ACT publishes. It tells us that that of those with a significant difference between HSGPA and ACT, a greater % of with higher GPA and lower test scores are female, under rep, or lower socio-economic.

ACT's own #discrepancy data shows us who benefits from testing.

A larger percent of girls outperform in the classroom, while a larger percent of boys outperform on the #ACT.

You know who largely doesn't benefit from ACT scores? poor students.

A much greater percent of students from lower income families have much higher HSGPA than they do ACT scores, for wealthy families its the reverse.

Who does the ACT and SAT help?

I sure wish that #edwriters didn't buy into the test publishers narratives so often.

ACT's and College Board's own data show us that there are very few people who are "saved" by the test.

the attached thread explores...

https://twitter.com/akilbello/status/1313545796219006976?s=20

this book is also seems like it would be an interestin read about the operations of the test companies.

https://twitter.com/akilbello/status/1458805958914809858?s=20

Seriously y'all.

The ACT/SAT ain't the magic you think it is.

Stats of ACT outperformers: 25 ACT 2.7 GPA
Stats of GPA outperformers: 17 ACT 3.6 GPA

Who would you rather in your classroom?

#EMChat

According to the College Board, relying on the SAT underpredicts HS gpa much more than relying on HSGPA overpredicts it.

Also note that in their research summary they mention the problem with only using HSGPA* but not the problems associated with using only SAT.

*strawman argument since no one has ever suggested doing this.

Holy crap! If I'm reading this right, colleges are biased toward those with high tests scores over those with high HSGPA, despite the fact that they persists at almost the same rate.

#DiscrepantScores

If you're interested in digging into the #discrepant reports, here you go:

ACT, 2014: slideplayer.com/slide/9394694/

CB 2010:
files.eric.ed.gov/fulltext/ED563…

ACT 2016: act.org/content/dam/ac…