Approaches to Parent Involvement for Improving the Academic Performance of Elementary School Age Children

19 randomized controlled trials (the review reports 19 RCTs meeting inclusion criteria)

Elementary-school-age children in parent-involvement intervention programmes

Campbell Collaboration systematic review restricted to studies with at least two groups and random assignment — the design screen almost nothing else in this literature passes. The trials are short (median 11 weeks) and small, and the outcome measures are not uniformly independent standardized tests, which is where the size of the pooled estimate should be discounted.

Pooled effect of parent-involvement programmes on academic performance d=0.45, from interventions with a MEDIAN LENGTH OF ONLY 11 WEEKS. The authors read this as compelling support for parent-involvement programmes as a supplementary intervention. Read against the rest of this archive, 0.45 from short small trials with mixed measure types is the classic efficacy-phase number that does not survive independent evaluation at scale — and See & Gorard 2015, screening the same field more harshly, found nothing of the sort.

Low by design — random assignment within families already homeschooling or schooling means the parents in both arms are equally selected.

Contradicted in spirit by See & Gorard 2015, which screened 4,898 reports and found no large robust evaluation and an almost even split between claimed successes and null or harmful results. Both cannot be right about the state of the field; the disagreement is about how much weight small short RCTs deserve.