Homeschooled Children's Social Skills

70 home-schooled children (32 boys, 38 girls, grades 3-6), compared against the 1,170 public-school children of the SSRS standardization sample

Home-schooled children in grades 3-6, all White, recruited as volunteers from two explicitly Christian home-school support groups in a single central Florida community; all had been home-schooled at least two consecutive years

This study exists to answer the one gap Medlin's own reviews keep flagging — almost every socialisation finding comes from the parent who chose home education, and "what homeschooled children think about their own social skills is less clear." So the measure here is the Social Skills Rating System Student Form, a standardized 34-item self-report yielding Cooperation, Assertiveness, Empathy, Self-Control and Total scores. That is a genuine methodological advance over parent rating, and it is undone by the comparison group: there is no contemporaneous control, only the SSRS 1990 standardization norms, so a 2006 convenience sample of all-White Christian homeschoolers is being compared to a 16-year-old national norm drawn from a differently constituted population. Nothing is matched on race, income, parental education or religiosity. The author himself notes the SSRS Student Form's weak psychometrics at the subscale level — alpha as low as .51 for Assertiveness, test-retest as low as .52, and parent-child agreement of only .03-.12. Grade D, and the design cannot support a causal reading.

Mean SSRS Total percentile ranks ranged from 55 (fifth-grade boys) to 94 (sixth-grade boys); every cell exceeded the 50th-percentile public-school average. Total scores were significantly higher than the norm group for fifth-grade girls and for sixth-grade boys and girls. Of 32 subscale cell means, 27 were above the standardization sample; the exceptions were third-grade girls' Self-Control, fourth-grade boys' Assertiveness and Self-Control, and fifth-grade boys' Assertiveness and Empathy. Significant subscale advantages appeared for fourth-grade girls (Empathy), fifth-grade girls (all four) and sixth-grade boys and girls (three of four). The pattern is: differences favour homeschooled children, grow with grade level, and are larger for girls. The correct reading is narrow — homeschooled children in this sample do not report themselves as socially deficient, which rules out one specific version of the isolation worry (that these children privately know they are struggling). It does not establish that home education produced better social skills, and Medlin's comparison design cannot distinguish the schooling from the family.

High and untreated. Volunteers from two Christian support groups in one town versus a 1990 national norm; the homeschool families are documented elsewhere in the paper as more educated and higher-income than the general population. Every route by which heritable parental traits could produce both the schooling choice and the child's social disposition is open.

Not independently replicated. It is directionally consistent with the parent-report SSRS study (Francis 1999) and with Smedley's 1992 Vineland results, but those share the same self-selection problem and are not independent tests. The one contrary datum in the corpus — Chatham-Carpenter's finding of lower rated closeness in peer friendships — comes from a different construct (relationship quality, not skill) and from behavioural diaries rather than a rating scale, so it neither confirms nor refutes this.