The Evidence on Teaching

A systematic review of the empirical research on selected aspects of homeschooling as a school choice

Ray BD · 2017

grade Dreviewdeveloper-lednot-applicable
Sample
narrative review; study counts not reported as a screened, quality-graded set
Population
US home-schooling research literature to 2017
Design
Author is president of the National Home Education Research Institute, the body that produces most of the primary studies being reviewed, and the review's own primary evidence is drawn from HSLDA/NHERI convenience samples. No design screen is applied: studies are counted by the direction of their result, not weighted by whether they had a comparison group.
Key findings
"The majority of peer-reviewed studies on academic achievement reveal a positive effect for the homeschooled students compared to institutional schooled students, while a few studies show mixed or negative results." Similar vote-counts are reported for social/emotional development and adult success. The review recommends future work use "well-controlled non-experimental designs" — an implicit concession that the reviewed base does not have them.
Genetic confound
High and unaddressed. The vote-count method treats a confounded positive result as evidence; under severe self-selection, a literature of confounded positives is exactly what a null causal effect would produce.
Replication notes
The vote-count findings are reproduced across Ray's own reviews (2013, 2017, 2026 update) drawing on an overlapping study set; that is repetition of a method, not independent replication of a result.

Effects

OutcomeMetricValueMeasureTimingVsHorizonClass
Academic achievement, home-schooled vs institutionally schooled (vote count)direction of result across studiesmajority positive, a few mixed or negative; no pooled estimate, no design screenstandardizedpooled across studiesactive-alternativeuncleardomain-skill
Social and emotional development (vote count)direction of result across studieslarge majority positivemixedpooled across studiesactive-alternativeunclearnon-cognitive

Cited by