Standards2030

The idea for the KMS originated in 2013, when the Next Generation Science Standards were published. I was dissatisfied with them for a number of reasons: science errors, lack of specificity, limited expression (they express only a fraction of the possible Performance Expectations.) and the air of arrogance about them. Those criticisms aside, the NGSS did show the value of expressing together both science content and what the student should do with the content. (The third strand, Cross-cutting Concepts, which I’ve described as Big Ideas elsewhere, I find not useful in the construction of learning goals.) They also exemplified separating writing the pieces—the DCIs, SEPs, and CCC—from composing the standards, as the DCI, SEP, and CCC first appeared in the earlier Framework (although there is some question about whether this work was truly independent.)

Prerequisite graphs go further back. Around 2005 I was at Project 2061 of AAAS, and work on the Atlas was in full swing. It’s interesting: part of my brain thinks that I conceived of prerequisite graphs independently, while working on the KMS, yet I can’t deny being exposed to them much earlier. Also while I was at Project 2061, I recognized the utility of expressing content ideas in tiny little pieces. At the time, this was because it was particularly useful for item writing.

Returning to the NGSS, while recognizing its contributions, I felt there was room for improvement. I immediately realized that standards must be expressed in a database. The model of combining entities in one column with entities from another column cries out for a relational database, and just as a matter of courtesy, it’s a little rude to hand someone a pdf and say “Here it is. Take it or leave it. And by the way, we’re going to lobby your legislators to make you take it.”

I set to work learning how to use a database and SQL. I used OpenOffice’s Base, later switching to LibreOffice. I actually did quite a bit of useful work back then, working within these databases’s self-contained user interfaces. But eventually I realized that a real product lives on the web. I picked up some HTML, CSS, PHP, and Javascript, but I never became proficient, and I didn’t have capital to hire someone proficient. For many years the project made only halting progress.

In the twelve years since the conception of the project, my interest has waxed and waned, but never fallen to zero.^[1] When in mid to late 2025, I realized that AI could compensate for my coding deficiencies, the project took off. It’s a wonderful feeling, to have had this vision in my head for so long, and now at last to have the ability to make it real.

Tom Regan

Thanksgiving weekend 2025