All verifications

Two RAND 2025 findings appeared in Episode 6's cold open in a way that merged them into a single shared concern. One finding is a parent statistic. One is a student statistic. They measure different worries and combining them misrepresents both.

What We Said

The cold open stated: "61% of parents and even half of high schoolers themselves, who are actively worried that using AI is going to harm their critical thinking skills."

What the Data Shows

The RAND 2025 nationally representative study (DOI: 10.7249/RRA4180-1, surveying middle and high school students, parents, teachers, and principals) produced two distinct findings.

Among parents: 61% reported concern that greater AI use will harm their child's critical thinking skills. That figure is accurately stated.

Among students: 51% of middle and high school students reported worrying about being falsely accused of AI cheating. That is what the 51% student figure measures.

RAND did not ask students whether they worry about AI harming their own critical thinking. The 51% captures anxiety about detection tools and false accusations -- which was the entire focus of Episode 4 -- not concern about cognitive harm.

What Was Conflated

The cold open presented the 51% student figure as if students share parents' concern about AI harming their critical thinking. The phrasing "half of high schoolers themselves, who are actively worried that using AI is going to harm their critical thinking skills" attributes the parent concern to students using the student false-accusation statistic. Those are two different things.

The distinction is not minor. Concern about AI harming one's thinking is a concern about cognitive development. Concern about being falsely accused of AI cheating is a concern about institutional fairness and consequences. A student worried about a false accusation may have no concern at all about cognitive harm -- or vice versa. Combining them into one sentence implies students have independently arrived at the same concern parents hold about critical thinking, which the data does not show.

The Corrected Framing

The finding that belongs with parents: 61% of parents worry that greater AI use will harm their child's critical thinking.

The finding that belongs with students: 51% of middle and high school students worry about being falsely accused of AI cheating.

Both are significant. Both come from the same RAND study. They should be presented as what they are -- two populations with two different concerns about AI in school -- not as a shared worry about the same thing.

How This Affects the Episode

The body of Episode 6 is accurate throughout. The Phung et al. research (102 students, single Python course, 8% of interactions involving evaluation, planful users achieving better outcomes) was correctly scoped and caveated. The Gonsalves distinction between critical thinking for the assignment and critical thinking toward the AI was accurately represented. The PAUSE framework introduction and the process protection framing were handled correctly.

The conflation appears only in the cold open and does not affect the episode's substantive conclusions or the accuracy of the PAUSE framework as presented.

---

Source: RAND Corporation, "Artificial Intelligence in K-12 Education," September 2025. DOI: 10.7249/RRA4180-1. Nationally representative sample of middle and high school students, parents, teachers, and principals.