Development in Context: Short-Term Trajectories of Expectancy and Values in Chemistry and Computer Science

• In Event: Students' Expectancies and Values After the Transition to Higher Education: A Situated Perspective

Abstract

Purpose and Theoretical Framework
Situated expectancy-value theory (Eccles & Wigfield, 2020) posits that expectancies for success and task values are key proximal predictors of educational outcomes, and that expectancy and value are not only individual characteristics but are also culturally and contextually bound. Task value is multifaceted: individuals value tasks because they are interesting/enjoyable (interest value), because the activity is important for one’s goals (utility value), or because the task is important to one’s identity (attainment value).
Research grounded in expectancy-value theory across several domains and levels of schooling indicates that students’ motivation tends to decline, on average, over time (e.g., Fredricks & Eccles, 2002; Kosovich et al., 2017; Robinson et al., 2019). However, little research has examined the potentially differential patterns and roles of changes in expectancies and the three types of value, or has focused on motivation trajectories among university students across different courses of study. A detailed understanding of when and how students are most likely to lose motivation is required for effective instructional design, particularly within the context of specific courses where teachers may have the opportunity to impact students’ motivation. Further, it is unknown whether overall patterns of change in motivation are reflective primarily of within-person developmental trends, contextual factors, or a combination of the two. Thus, in the present study we addressed the following research questions:
1. How do students’ expectancy for success and task values change within a single semester?
2. Do motivation trajectories differ across different subjects of study?

Methods
Students in computer science and chemistry courses (N=2,101) completed four surveys during the Fall 2020 semester. Measures (αs=.83-.95) included self-efficacy as an indicator of expectancy for success, adapted from Midgley et al. (2000), along with utility, interest, and attainment value scales adapted from Conley (2012).

Results
Following confirmatory factor analyses and measurement invariance testing, latent change score analyses (McArdle, 2009; Table 1) indicated significant declines for all four motivation constructs during the beginning of the semester, with patterns of stability near the end of the semester (Figure 1). Differences across chemistry and computer science were unique to each motivation construct and timepoint; for example, attainment value showed no significant differences across domains, expectancy for success showed differential change patterns from T1 to T2 and from T3 to T4, whereas utility value trajectories differed across domains only between T2 and T3.

Discussion
In alignment with prior research, we found overall declines in motivation (Jacobs et al., 2002) largely occurring at the beginning of the semester (Benden & Lauermann, 2020). Results suggested an overall similar shape to changes across constructs and contexts. However, trajectories were not uniform across computer science and chemistry learners, contributing unique information about the contextual nature and timing of motivational development across different domains of study and for specific motivation constructs. Understanding what developmental patterns are common across settings and which are more particular to specific settings and students adds essential understanding of motivational strengths and needs in real-world classroom contexts, with implications for the design and timing of interventions.

Authors

• Kristy A. Robinson, McGill University

• So Yeon Lee, McGill University