The Evidence on Teaching

Abused Homeschooled Children Deserve Legal Protections: A Response to Ray and Shakeel (2022)

Stewart G, McCracken C · 2023

grade Dcritiquedeveloper-involvednot-applicable
Sample
not applicable — a 4-page commentary, no original data
Population
not applicable
Design
A short published response, not a study. Authors are from the Coalition for Responsible Home Education, the advocacy organisation campaigning for homeschool regulation — the mirror image of NHERI's position, and the reason this file is flagged as not independent. The archive's `independence` vocabulary has no slot for "opponent-led"; `developer-involved` records the load-bearing fact, which is that the authors have a declared stake in the answer. Recorded because the adversarial pass on Ray & Shakeel 2022 should be visible in the corpus rather than left to a reader to go find, and because a critique whose own bias runs the opposite way is exactly what a reader needs alongside a developer-led primary study.
Key findings
Two methodological objections plus one about motive. First, the sampling: snowball recruitment through an online panel overrepresents certain demographics, so the "nationally representative" claim is overstated. Second, the instrument: the survey's abuse items are short and were not properly validated. Third, and this is advocacy rather than methodology, the authors argue the study's purpose is to forestall policy measures that would protect homeschooled children. The authors concede that Ray and Shakeel's design is novel and valuable — that concession is worth noting, because it means the two poles of this debate agree that a comparison-group survey is the right instrument and disagree about this execution of it. Neither side has produced a version that a disinterested party ran.
Genetic confound
not applicable
Replication notes
not applicable — a single commentary. The substantive dispute it opens has never been resolved by an independent re-run of the comparison.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Validity of Ray & Shakeel's null finding on homeschool maltreatmentqualitative critiquesampling (snowball, demographic overrepresentation) and instrument validation judged inadequate to support the conclusion; design concept accepted as valuablemethodological critiquenot applicablenonenot-applicablehealth

Cited by