Effects of embedded morphological instruction on children's morpho-syntactic production

Thesis (Ph.D.)--University of Washington, 2017-06Being able to use the words that are learned and encountered in school is an important skill for communicating and displaying understanding. Students are increasingly expected to be able to appropriately and effectively use complex academic words in their writing, yet we know little about this productive aspect of word knowledge. This study explores the effects of relevant word characteristics, student characteristics, and an embedded morphologically-focused instructional intervention for 499 fourth and fifth grade students randomly assigned to treatment and control classrooms (n = 233 and 266, respectively) on their morpho-syntactic production. Considering word-level SFI, word-level phonological and orthographic transparency (for Morphological Production outcome only), student-level comprehension and morphological knowledge at pretest, and classroom level instructional condition (treatment or business-as-usual control), item-level responses were predicted using cross-classified mixed models with random effects for two more-or-less constrained measures of morpho-syntactic production: Morphological Production and Sentence Combining Derivations. Results suggest that instruction has a unique and positive effect on students’ control of morpho-syntactic word forms produced in sentence-level syntax activities. Students’ pretest comprehension, as a proxy for general literacy skill, and students’ pretest morphological skill were both uniquely predictive of both outcomes, for instructed and non-instructed words. Word-level frequency (SFI) was predictive only for deriving non-instructed words on sentence combining, while phonological and orthographic shift type was predictive only for deriving instructed words on morphological production. Implications for word-level instruction are discussed