Paper Summary
Share...

Direct link:

Core Academic Language Skills: A Relevant Construct to Understand and Promote Adolescents' Expository Writing Quality?

Sat, April 9, 8:15 to 9:45am, Convention Center, Floor: Level One, Room 153

Abstract

Objectives and Theoretical framework
Academic language proficiency has been a topic of interest given its hypothesized contribution to skilled reading for adolescents (Cummins, 2001; Nagy & Townsend, 2012; Valdés, 2004). Indeed, studies using the recently operationalized construct of Core Academic Language Skills (Uccelli et al., 2014; Uccelli et al., 2015), which seeks to explore a broader constellation of language skills that support academic text comprehension (managing densely packed syntax, understanding precise academic language and language that signals how ideas and propositions are connected within a text, etc.), finds that CALS uniquely predict adolescents’ reading comprehension above and beyond the variance contributed by academic vocabulary, word reading skills, and socio-economic status. When reading and writing, students face common challenges, e.g., understanding and producing precise language, densely packed structures, logical connectives (Bailey, 2007; Schleppegrell, 2004; Snow & Uccelli, 2009). Thus, in this study, we extend the CALS construct to the domain of writing. Specifically, we explore CALS as evidenced in students' expository writing and examine the association between essay-based CALS and expository writing quality.

Methods
A total of 244 4th-to-6th grade students from seven schools in two large urban school districts in the Northeastern U.S. participated in this study. The sample was approximately balanced by gender and socio-economic status (SES). Most students were classified by the school as English proficient and only 4% as English Learners.
Four assessments were administered:
(1) CALS-Instrument | (44 items, α = .93; split reliability=.90): group-administered test that evaluates receptive CALS in grades 4 to 6. Tasks assess a range of skills through multiple choice, matching, or short written responses. Form 1 was used for this study and included six tasks: Connecting Ideas, Tracking Themes, Organizing Texts, Breaking Words, Comprehending Sentences, Identifying Definitions.
(2) Productive Essay-based CALS | Students’ written expository texts were coded for linguistic features in order to generate domain-specific scores for lexical and syntactic packing, text connectivity, and genre features. A number of indices were generated using CohMetrix, the most comprehensive automated text analysis tool available presently (McNamara et al., 2005). Confirmatory factor analysis was used to generate domain-specific student level scores.
(3) Writing quality | Essays were also be scored for overall writing quality using the 6-point-scale NAEP Scoring Guide-To Inform (NAEP, 2011) by experienced ELA teachers who were blind to the research goals and design. This rubric evaluates four aspects of writing (1) Development of ideas, (2) Organization, (3) Sentences/Word Choice and (4) Genre Elements.

Results
Mirroring analysis conducted on the CALS-Instrument, productive CALS scores loaded onto a single factor and showed variability within and across grades. Preliminary correlational analysis suggests that students’ receptive CALS as measured by the CALS-I are strongly correlated with their productive CALS scores (0.79). Analysis is presently on-going and will examine the relationship between productive CALS scores and writing quality.

Significance
This study examines the potential of a common underlying CALS construct that might support both text comprehension and production.

Authors