This website uses cookies

Read our Privacy policy and Terms of use for more information.

In 1975, the British economist Charles Goodhart made an observation about monetary policy known today as Goodhart’s law, which states, "When a measure becomes a target, it ceases to be a good measure."

Education may provide some of the best examples of what Goodhart meant.

Test scores turn into measures of learning. Time spent in seats turns into measures of learning progress. A diploma turns into a measure of readiness. A list of summer reading turns into proof that students participated in summer reading. The proxy starts as a practical way to see something difficult to see; eventually, we forget it was ever a proxy.

Artificial intelligence only makes it harder to ignore. Students can now submit well-written essays, completed problem sets, or professional presentations without doing the thinking those products once signaled. That is the starting point for a new white paper from the Stanford Accelerator for Learning and ETS. Traditional assessments might tell us whether a student arrived at the right answer, Cal State Chancellor Millie García says in the paper, but they tell us little about how the student got there. The authors of the paper call for a transition from tests to collecting evidence: portfolios, conversations, performances, revision, and other ways of documenting student thinking, creativity, collaboration, and progress.

Several pieces in today's reading list raise versions of the same problem:

  • In Education Week, teachers say mandatory summer reading produces cramming and resistance and not actual reading. AI-written summaries of books have only made it worse.

  • In Education Next, Liz Cohen defends calculus against people claiming it is an outdated proxy for college readiness while acknowledging that, for many selective universities, it is a proxy indeed.

But today's reading also contains a warning against confusing imperfect measures with worthless ones.

  • In The 74, Conor Williams makes the argument that standardized tests and transparent goals are necessary for rebuilding trust in public schools. Without data, families and policymakers are left with anecdotes.

  • And, as K-12 Dive reports, districts with persistently high absenteeism were about 50 percent further from recovering in math than districts that never experienced high absenteeism. The right measure can uncover important insights.

So the question isn't whether to measure. It's whether we are measuring the thing we actually care about, and what happens once everyone learns how to satisfy the metric. AI didn't create that problem; it just made it hard to miss.

K-12 Education

Higher Education

Federal Policy & Politics

Early Learning & Child Care

State & Local News

Educator Talent & Staffing

AI & Technology

Student Health, Safety & Nutrition

Workforce & Career Pathways

School Choice

Also Reading

Job Opportunities

Looking for your next opportunity in education? W/A Jobs features 3,500+ career opportunities from 300+ organizations across the education industry.

Whiteboard Notes

Whiteboard Notes

Stay informed with our weekly newsletter recapping the major education policy and industry news. Published every Friday, Whiteboard Notes is trusted by thousands of education leaders, investors, gr...

Recommended for you