Search

There is increasing interest in the role of phonological awareness across languages. Research is uncovering cross-language effects of phonological awareness upon English reading, even from nonalphabetic languages. However, little of this research has focused on examining the extent to which multiple measures of phonological awareness indicate a single construct within or across languages. This article updates 2 recent reviews of the literature by fitting rival a priori models of multiple measures in order to test within-language and across-language structure among multiple phonological awareness tasks. Although the number and types of languages covered were quite limited, the results demonstrate high cross-language correlations, suggesting that measurement error has attenuated prior estimates of the cross-language correlation of phonological tasks. The current results suggest that in alphabetic languages, there is empirical support for phonological awareness as a unitary ability within English and other languages. In Korean and in Spanish, phonological awareness may operate as a language-general construct. In Cantonese and Mandarin, the results were less clear. The results also highlight the limitations of the current research base and important areas for future investigation.

There is increasing interest in the role of phonological awareness across languages. However, the role of phonological awareness in various languages may differ by features of the languages as well as by features of the speakers. The current meta-analysis catalogs these relations and examines factors that may have influenced how closely related these measures are between English and other languages. Studies were selected reporting cross-language correlations with English phonological measures involving children in preschool to upper elementary (ages 4-10 years). A random effects model of 101 correlations within 38 studies suggested that the cross-language correlation of phonological tasks is influenced heavily by the language used and to some extent by the linguistic grain size of the tasks used. Although no strong effect of age was found, variability across studies was significant, reflecting unexplained differences likely due to features unable to be modeled in the current sample. The present study suggests limitations in the current research base for drawing cross-language inferences. (Contains 2 tables, 3 figures and 1 footnote.)