Validity of large-scale reading tests: A phenotypic and behaviour-genetic analysis.

Each year, all Australian students in grades 3, 5, 7, and 9 sit nationwide large-scale tests in literacy and numeracy, which have their validity frequently questioned. We compared the performance of Grade 3 twins on these large-scale reading tests with their performance on three individually administered literacy tests in comprehension, word reading and vocabulary within a genetically sensitive design. Comprehension, word reading, and vocabulary accounted for a substantial amount of the variance in school reading tests. Performance on large-scale reading tests and individually administered tests was moderately to substantially heritable and the same genes contributed to performance in both types of test. These results confirm that large-scale school reading tests measure, at least in part, the literacy skills tapped by individual tests that are frequently considered to be the "gold-standard" in testing.