CMO/Advisor/Tsundoku ex-@ga @aplaceformom @expedia @mckinsey @wharton @proctergamble Author- MarketingBS
Seattle, WA
Joined April 2010
- Tweets14.9K
- Following783
- Followers41.5K
- Likes202K
🐂 Edward Nevraumont 🐂 retweeted
You’ve probably used AI for work today.
Now picture an 11-year-old using it to finish her core schoolwork in two hours a day.
That’s Everest Nevraumont’s routine at @Alphaschool in Austin, and she’s given a @TEDx talk on how she does it.
She’s coming with her dad, @Ednever, Senior Advisor at @warburgpincus and former CMO of @GA_TORONTO, to talk about what AI is really doing for kids’ learning and where it’s mostly hype.
If you build in edtech, shape AI policy at a school, or have kids of your own, come hear it in person and bring your questions.
Get your early-bird ticket: luma.com/future-of-education…
Buchanan was clearly a gay president (likely had an affair with William Rufus, VP of Pierce). But he was also one of the worst presidents so no one wants to claim him. Instead we call him our only “bachelor president”
This is clearly the right answer.
The issue was that he was also a terrible president so no one wants to claim him….
You actually don’t. NWEA has shown there is no impact of income (or demographics) on MAP growth once you control for MAP baseline
If an org can show faster than average MAP growth that is real regardless of selection effects
“While absolute achievement scores (a student's fixed baseline percentile) are heavily influenced by household income, the rate of change in that score over time is decoupled from socioeconomic status.”
Yes. And that is why high income kids have higher MAP scores and why higher MAP score kids improve their MAP scores faster than lower score kids
But once you control for MAP income has barely any impact in the rate of change of MAP %iles
Study: nwea.org/poverty-and-school-…
Fusion: “students can learn material at double the normal pace”
Acton: “Most Acton students progress through their core curriculum at two to three grade levels in a single year”
Khan: “students master concepts years ahead of their traditional grade peers”
Alpha isn’t selective. If a kid can focus for about 20 minutes they are accepted.
Why do you think demographics are a better comparison group than MAP scores?
So 170 is now about 90th percentile
And 180 is still 99.9th percentile?
Seems like it makes it hard to make a granular decision at the top end.
7.3mm take them per year. About 20%
But the point isn’t MAP, it’s improvement on MAP based on the same kids taking the MAP in the same %ile three months earlier
I’m honestly not sure how you measure better
Most don’t make learning claims they just infer them by listing things like SAT scores, Ivy matriculation, selectivity, doctorate rates of teachers, college courses offered, and AP pass rate. All driven by selection effects
Alpha just claims 2.8x on MAP scores
I paid higher tuition at the last school the kids were at. We only pay $25k at one of the less fancy schools. I have every incentive to hold them to a high standard or I will pull them out of here too
The overall results are released. The percentiles are real. We just don’t know which schools are doing better than average given starting points
Alpha definitely missed their target in AZ, but they still improved significantly better than the average given starting point
Meanwhile all the schools that are NOT being criticized don’t share anything at all….
Alpha has a lot of things it needs to work on but at least I can see how my kids are doing and work through the gaps that happen. Other schools we have been at are complete black box
The claims are just that kids improve MAP scores 2.8x faster than other kids of the same MAP scores across the US. That is as apples to apples as it gets
AZ school did not hit the internal goals but the kids in the program still improved their MAP scores significantly more than the rest of the state (they were still lower than average at the end of the year, but they started WAY lower than average)
I think they are convinced it works and they are “hell bent” on expanding it to meet demand. They’ve launched ~50 schools in the last year. I expect operationally they are strapped. Why would they limit demand to run more tests when it won’t convince critics anyway?