Predicting Profile Permanence: When Is Motivation Stable, Why Does It Change, and What Are the Consequences?

• In Event: Person-Centered Research: A Methodological Answer to Motivation Theory's Most Critical Questions

Abstract

Profile-centered motivation research primarily consists of single-timepoint designs and consequently has not considered profile stability or predictors of profile shifts. This study investigated exam achievement as a predictor of profile shifts to capture the dynamic nature of students’ motivation. We adopted an integrative motivational perspective, drawing from Achievement Goal Theory (Ames, 1992) and Expectancy-Value Theory (Wigfield & Eccles, 2002) to holistically describe students’ motivational orientations.

College anatomy students (n = 306) reported their motivation (value, perceived competence, and achievement goals: mastery, performance-approach, performance-avoidance) at the beginning, middle, and end of a semester. Using i-states-as-objects analysis with hierarchical and nonhierarchical clustering procedures, five profiles were identified: highly-motivated-by-any-means (high on all motivational variables; nTime1,Time2,Time3 = 58, 51, 31), intrinsically-motivated-and-confident-with-average-performance-goals (high mastery goals/competence/value, average performance goals; nTime1,Time2,Time3 = 102, 60, 61), intrinsically-motivated-and-confident-with-low-performance-goals (high mastery goals/competence/value, low performance goals; nTime1,Time2,Time3 = 48, 53, 39), performance focused (low mastery goals/competence/value, high performance goals; nTime1,Time2,Time3 = 63, 73, 72), and amotivated (low on all motivational variables; nTime1,Time2,Time3 = 28, 49, 58). All but the intrinsically-motivated-and-confident-with-average-performance-goals profile mirrored similar studies (Linnenbrink-Garcia et al., 2012, 2014; Wormington et al., 2014), suggesting profile replicability. The first three profiles reported equally high behavioral engagement and course achievement, while amotivated and performance-focused profiles were less adaptive for these outcomes.

Results examining profile shifts suggested substantial profile instability, with 40-50% of participants shifting profiles at each timepoint. Unsurprisingly, students were unlikely to completely switch motivational orientations (e.g., highly-motivated-by-any-means to amotivated). The amotivated profile was the most stable at both timepoints, with students unlikely to shift into more adaptive profiles. Conversely, the intrinsically-motivated-and-confident-with-average-performance-goals profile was the least stable, with students often maintaining high mastery goals/competence/value but either increasing (shift to highly-motivated-by-any-means profile) or decreasing their performance orientation (shift to intrinsically-motivated-and-confident-with-low-performance-goals profile).

Finally, we used logistic regression to examine whether participants’ mid-semester exam score predicted profile shifts from the middle to end of the semester and if that association differed by mid-semester profile. Main effects were qualified by significant interactions for the highly-motivated-by-any-means and intrinsically-motivated-and-confident-with-low-performance-goals profiles. As exam scores for students in these profiles increased, the probability of shifting profiles decreased. The opposite pattern was observed for students in the amotivated profile; as their exam scores increased, the probability of shifting profiles increased.

Results provide insight into when and why student motivation changes, distinguishing temporary from longer-lasting motivational loss. The amotivated profile appeared particularly concerning given high stability and maladaptive outcomes; however, high exam achievement, though relatively rare for students in this profile, may spur motivational change among amotivated students. Results also underscore unique insights gained from profile-centered findings with respect to profile stability; amotivated and performance-focused profiles, though endorsing similar mastery goals/competence/value, displayed very different overall stability and profile shifts. Finally, results address theoretical questions, such as the performance-approach goal debate: profiles with high mastery goals but varying levels of performance goal endorsement, for example, were equally adaptive, suggesting no benefits or drawbacks to performance goal endorsement Performance goal endorsement without accompanying mastery goals, value, and competence, however, appears to be maladaptive.

Authors

• Stephanie V. Wormington, Michigan State University

• You-kyung Lee, Michigan State University

• Christopher Seals, Michigan State University

• Joshua Michael Rosenberg, Michigan State University

• William Saltarelli, Central Michigan University

• Cary J. Roseth, Michigan State University

• Lisa Linnenbrink-Garcia, Michigan State University