Homeschooling and the Question of Socialization Revisited

narrative review of small comparative and survey studies

Home-schooled children and adults in the US

Vote-counting narrative review. The underlying corpus is almost entirely small, self-selected, parent-reported or self-reported comparisons; no included study has random or quasi-random assignment, and the review does not weight studies by design.

Home-schooling parents report expecting their children to get along with people of diverse backgrounds and providing social opportunities outside the family, and rate their children's social skills as at least as good as other children's. Compared to conventionally schooled children, home-schooled children are reported to have higher-quality friendships and better relationships with parents and other adults, moral reasoning at least as advanced, less emotional turmoil and fewer problem behaviours in adolescence, and civic engagement in adulthood. Medlin's own hedge is the most useful line in the paper: "What homeschooled children think about their own social skills is less clear." He concludes that an alarmist view is not supported by the research.

High and essentially untreated. Almost every outcome is reported by the parent who chose home-schooling, and the comparison children come from different families.

The favourable direction is reproduced across Medlin's own 2000 and 2013 reviews and Ray's reviews from an overlapping corpus. The independent surveys (Kunzman & Gaither) read the same corpus and decline to draw the conclusion, which is a disagreement about evidentiary standards rather than about findings.