A Meta-Analysis of Morphological Interventions in English: Effects on Literacy Outcomes for School-Age Children

Abstract

The purpose of this study is to examine the effects of morphological instruction on language and literacy outcomes by synthesizing 92 standardized mean differences (d) from 30 independent studies. Findings show a moderate overall effect of morphological instruction ([dbar] = 0.32), suggesting that children receiving morphological instruction performed significantly better on measures of literacy achievement than comparison groups. Moderator analyses showed that intervention effect varied depending on the literacy outcome. There were significant and moderate intervention effects on morphological knowledge ([dbar] = 0.44), phonological awareness ([dbar] = 0.48), vocabulary ([dbar] = 0.34), decoding ([dbar] = 0.59), and spelling ([dbar] = 0.30) but not on reading comprehension or fluency. Results also suggested differences in effectiveness related to age and research design but not unit of intervention, scope, length, and learner type. Effect sizes decrease by school level (e.g., greater for younger students than middle school and upper elementary students). Also, there were larger effects for quasi-experimental than experimental studies and for researcher-designed measures than for standardized measures. Implications for educational settings and research agendas are discussed.

ACKNOWLEDGMENTS

We thank Dr. Terezinha Nunes, Dr. Michael Kieffer, and Dr. Nonie Lesaux for their prompt responses and feedback. Their data have improved this analysis. We also thank the anonymous reviewers and Joanne Carlisle for their helpful feedback

Notes

¹Authors attended several educational and reading conferences such as the American Educational Research Association, Council for Exceptional Children Convention & Expo, Society for the Scientific Study of Reading, Institute for Educational Sciences Annual Conference, and the Literacy Research Association and polled experts to find any studies that were never published or upcoming unpublished studies related to morphological instruction.

²When studies reported incomplete data, the primary authors were contacted repeatedly via e-mail, via phone, and at conferences to obtain the information to calculate effect sizes. Researchers were sent requests for information on the intervention related to moderator variables as well.

³Time frame of studies published since 1980 was chosen based on the history of the study of morphological knowledge in relation to literacy. For example, in the 1970s, researchers seemed focused on examining the correlations between morphological knowledge and general reading measures (Brittain, 1970), whereas by the 1980s, researchers determined the frequency of morphologically complex words in text (Nagy & Anderson, 1984) and began to study morphological interventions (White, Sowell, & Yanagihara, 1989; Wysocki & Jenkins, 1987).

⁴Studies in which reading instruction did not occur in English were excluded, eliminating three studies with instruction in English and Spanish and seven studies with instruction in languages other than English (Arnbak & Elbro, 2000; Cañado, 2006; Carlo et al., 2004; Casalis & Cole, 2009; Chow, McBride-Chang, Cheung, & Chow, 2008; Elbro & Arnbak, 1996; Lyster, 2002; Morin, 2003; Packard et al., 2006; Tomesen & Aarnoutse, 1998). In addition, three single subject studies (i.e., Carlisle, 2007; Edwards, 1982; Katz & Carlisle, 2009) and studies with no control group (Hesse, Robinson, & Rankin, 1983; Long & Rule, 2004; Tsesmeli & Seymour, 2009; White et al., 1989; Wysocki & Jenkins, 1987) were excluded.

⁵We eliminated two effect sizes (one from Nunes, Bryant, & Olsson, 2003, and the other from the second study in Bryant et al., 2006) bigger than |2.00|, as we considered them outliers, which would threaten the normality of effect-size distribution (Viechtbauer & Cheung, 2010).