Language in science performance: do good readers perform better?

Abstract

Science performance is highly affected by students' reading comprehension. Recently, there has been a growing attention to the role of linguistic features for science performance, but findings are ambivalent, especially when looking into item word count. The aim of this study was to investigate the interaction of students' reading comprehension and item word count of given science measures on performance, controlling for students' cognitive abilities, gender, family language, and school track. The sample consisted of N = 2051 German students in grades 10 and 11. Students completed (scientific) literacy measures. We then applied a multilevel logistic regression to investigate the hypothesized interaction effect of reading comprehension and word count on students' science performance. The results showed a significant interaction of students' reading comprehension and word count on science performance, controlling for several covariates. Particularly students with high reading comprehension benefit from science items with increasing word count. Our findings empirically support previous research, showing that reading comprehension is crucial for science performance and enhances the interaction between reading comprehension and linguistic features of written text in science subjects. Finally, theoretical and practical implications and limitations of this study are discussed.

What skills related to the control-of-variables strategy need to be taught, and who gains most? Differential effects of a training intervention.

“…In line with theoretical assumptions and empirical findings of previous studies (Chen & Klahr, 1999; Cruz Neri et al, 2021; Wagensveld et al, 2015), we expect children’s (a) prior knowledge (i.e., their CVS skills measured at pretest), (b) reasoning abilities, and (c) reading comprehension to explain variation in differences between the intervention and the control group. Building on the large sample size of our study (over 600 learners), we can examine whether these variables interact with the effects of the two conditions for the overall CVS score, as well as for the four more specific CVS subskills.…”

“…Regarding reading comprehension, we undertook careful pilot testing with our CVS test, including cognitive interviews, to ensure that students' achievement did not depend on their reading comprehension to an unduly extent. Nevertheless, reading comprehension is substantially related to more general verbal abilities (Cain et al, 2004;Schroeder, 2011) which have been shown to be predictive of children's development of skills related to the CVS (van der Graaf et al, 2016) as well as to science learning (Cruz Neri et al, 2021). The better the children's verbal abilities, the better they might be able to follow the instructors' explanations, and engage in more effective verbalization as well as verbal interactions with their peers during collaborative inquiry (Van Boxtel et al, 2000).…”

“…The better the children's verbal abilities, the better they might be able to follow the instructors' explanations, and engage in more effective verbalization as well as verbal interactions with their peers during collaborative inquiry (Van Boxtel et al, 2000). Better comprehension also helps integrating sentences into coherent mental representations, in turn fostering students' science literacy (Cruz Neri et al, 2021;Hall & Miro, 2016).…”

LATIC–A linguistic analyzer for text and item characteristics

“…In terms of scientific aspects, many authors traditionally tend to manually tag, code and count text and item characteristics, such as word count [ 9 ] or prepositions [ 19 ]. While doing this, the human raters might fall back on several different tools, which was the case in our study.…”

“…The analyses of linguistic text and item characteristics might not only be necessary for investigating research questions regarding language in texts and items, but also for practitioners, such as teachers and educators. Traditionally, texts and items are analyzed manually [ 9 , 19 ], which is very time consuming. LATIC is a user friendly and free to use java application enabling users to analyze texts and items in four languages (English, French, German, and Spanish) regarding their linguistic characteristics.…”