Does spelling instruction make students better spellers, readers, and writers? A meta-analytic review

53 studies, 6,037 students, 58 effect sizes

Kindergarten through Grade 12 (roughly ages 5-18)

Comprehensive meta-analysis of experimental and quasi-experimental spelling studies (Reading and Writing 27(9), 1703-1743), answering eight pre-specified questions about the effect of formally teaching spelling on spelling, phonological awareness, reading and writing. The design-class mix is the usual one for this literature, so grade C. The load-bearing feature is that it separates the outcomes rather than reporting a single "literacy" effect - which is what allows it to report a transfer and a non-transfer in the same paper.

Formal spelling instruction versus no spelling instruction or unrelated instruction: ES = 0.54 (p <= 0.001). MORE formal instruction versus LESS formal instruction: ES = 0.70 (p < 0.01) - i.e. within spelling teaching, more explicit is better than less explicit, which is the same direction the archive already records for reading. Formal spelling instruction improved students' spelling, their phonological awareness, and their READING skills. But it "did not translate into better performance on extended writing tasks". That asymmetry is the finding: spelling transfers DOWN into the code (phonological awareness, word reading) and does NOT transfer UP into composition. Anyone selling spelling instruction as a writing intervention is selling the one thing this meta- analysis explicitly failed to find.

Low for direction (experimental and quasi-experimental comparisons). Medium for magnitude, given the mixed design quality and the usual reliance on researcher-administered spelling measures, which are close to the taught content.