Family effects inflate polygenic IQ prediction

PGS predict intelligence and school outcomes much more strongly across unrelated people than they do between siblings; the within‑family (dizygotic twin difference) estimates are substantially smaller, indicating that ancestral, mating, or shared‑family processes amplify population‑level PGS signals. That gap is measurable across childhood to early adulthood in the UK Twins Early Development Study sample. — If PGS effects are partly family‑level rather than individual, policymakers and journalists should be cautious about using PGS as indicators of individual ability or as justification for interventions like embryo selection or educational sorting.