The Complexities and Challenges of Understanding Individual Differences and the Development of Multiple Dimensions of Writing Self-Efficacy

• In Event: 61.015 - Embracing and Modeling the Complex Dynamics of Motivation and Engagement: Contextual, Temporal, Dynamic, and Systematic

Abstract

Purpose & Theoretical Framework
Effective writing often requires a tremendous amount of time, energy, cognitive effort, perseverance, and, for many, emotional resiliency. In particular, self-efficacy, the confidence individuals have for performing successfully in a given domain (Bandura, 1997), is a critical factor in writing performance. Decades of research show that writing self-efficacy strongly predicts students’ effort, persistence, and ultimately, their performance on writing tasks (Bruning et al., 2013; Pajares, & Valiante, 1997). Illustrating the complexity of writing self-efficacy, recent research suggests that students hold separable efficacy beliefs about writing ideation, mechanics/conventions, and self-regulation (Bruning et al., 2013; De Smedt, Merchie, Barendse, Rosseel, De Naeghel, & Van Keer, 2018). However, no longitudinal investigations of writing self-efficacy exist. Such work is necessary to understand the development of students’ efficacy for writing across time. The current study used an accelerated growth model to understand individual differences and the development of three latent factors of student writing self-efficacy (ideation, mechanics, and self-regulation) across grades 3-11. Research questions that guided this work were: (1) What are the developmental trajectories of students’ self-efficacy for writing ideation, conventions, and self-regulation between grades 3-11, and to what extent are these developmental trajectories related? and (2) Do individual characteristics such as gender, ethnicity, or socioeconomic status moderate these trajectories?
Methods
Participants in this study included approximately 1,700 students between grades 3 - 11 enrolled in English Language Arts classes in a suburban public school system. Students completed a self-report survey twice annually (during the fall and spring) over three years for a total of six measurement occasions. Data were collected from seven cohorts of students using a cohort-sequential design, allowing for the approximation of a more traditional longitudinal panel design.
Data Analysis
We used covarying second-order latent growth curves (e.g. Hancock, Lawrence, & Kuo, 2001) to examine the unique developmental trajectories of each of the three writing self-efficacy factors (ideation, mechanics, and self-regulation) as well as relations between features of these developmental trajectories. Incorporating discontinuities in the growth models at transitions between elementary and middle school as well as between middle and high school allowed us to test the extent to which growth in writing self-efficacy differed across developmental periods. Preliminary findings suggest different trajectories for different groups of students as well as quadratic effects over waves of data.
Results & Significance
Initial models demonstrated different trajectories for different groups of students as well as quadratic effects over waves of data. We will elaborate on these findings in the full presentation. Understanding the development of multiple dimensions of writing self-efficacy over time is inherently complex and the challenges that accompany such complexity warrant discussion. Appropriately handling missing data is one of the primary challenges of this project due to the need to consider potential cohort differences and the impact of school feeder patterns on student attrition.
This work contributes to our growing understanding of the development of multiple factors of student writing self-efficacy across developmental levels, which can lead to instructional practices fostering students’ writing performance.

Authors

• Sharon Zumbrunn, Virginia Commonwealth University

• Michael Broda, Virginia Commonwealth University

• Sarah Anne Marrs, Virginia Commonwealth University

• Eric Ekholm, Virginia Commonwealth University