Effect of Prior Knowledge on Good and Poor Readers' Memory of Text (the 'baseball study')

64 US 7th-8th graders / 1 custom text

64 junior-high students crossed on reading ability x baseball knowledge.

2x2 observational design measuring whether pre-existing domain knowledge substitutes for reading skill on a single researcher-designed baseball passage. Not an intervention.

The famous demonstration that domain knowledge can outweigh reading-ability differences on comprehension of a topic-matched text. Heavily over-cited: N=64, single text, unreplicated, restricted range. Shows knowledge helps comprehension WITHIN a known domain (near transfer); does NOT show teaching knowledge raises general comprehension.

Severe: both baseball knowledge and reading ability are selected-into/heritable; the 'effect' is confounded by whatever produced the knowledge.

Not replicated in 30+ years; the key ability x knowledge interaction may be non-significant (Seidenberg via Shanahan). An existence proof, not a robust effect.