UC Irvine

This dissertation is guided by my driving research question: what works for whom in education and why? I examine issues of instruction, assessment, and research approaches that begin to answer this question. Study 1 examines a classroom intervention’s child x instruction interactions. Study 2 examines underlying cognitive factors, including non-verbal cognitive factors, of morphosyntactic performance on different language tasks. Study 3 highlights research opportunities for better understanding what works for whom and why by identifying and analyzing substantive and methodological themes in highly cited articles from research on math and language and literacy learning. Study 1: I investigated child × instruction effects on mathematics gains in first grade within a large randomized controlled trial. Results revealed that initially lower scoring children performed better without the treatment of Math PALS. Implications include the possibility that children with lower initial skills may need more time in small groups with their teacher to learn math skills, versus more time with peers. For children with higher skills, this study suggests working with peers is educationally beneficial for their learning outcomes as measured on standardized assessments. Study 2: In Study 2, I examine assessment tasks in light of what works for whom and why with an attention to domain general cognitive factors. Differences in the tapping of domain general non-verbal factors on language performance tasks are explored in a series of four hierarchical linear regressions predicting cloze and narrative performance in English and Spanish. The results reveal the differential contribution of nonverbal cognitive skills across English and Spanish and by task. The findings suggest cognitive demands vary for bilinguals based on the language of assessment and the task. This indicates that non-verbal IQ domain general cognitive factors (processing speed and working memory) support how students perform on what are often considered more domain specific tasks, such as my study’s grammatical tasks. This work has implications for understanding what kinds of instruction and assessment might work better for bilingual students with different levels of English exposure. Study 3: In Study 3, I compare mathematical cognition and learning research to reading research, a partially overlapping field with important similarities in research questions, methods, and populations of study. In this systematic literature search of highly cited papers in both areas, themes across the fields emerged: a reliance on longitudinal panel data, a focus on children in the early school years, and a lack of interventions shown to affect students’ long-term success (including older students’ achievement). While themes coded by decade reveal progress in approaches over time, some challenges still persist to solve. Such challenges arise from thematic constraints across fields such as an overreliance on longitudinal panel data versus experiments. This big picture approach to characterizing the field is potentially useful for improving our understanding of what works for whom and why in future work. In chapter 5, I discuss the implications of these findings for educational research and practice. In particular, I suggest that approaches of instruction, assessment and research needs more nuance in its approaches and vantage points. This is both at the level of the child and the level of the research questions that drive findings.