Overcoming Confusion: An Intervention to Regulate Confusion During Complex Mathematics Problem-Solving

• In Event: Emotions in Student and Teacher Learning: New Insights and Educational Applications

Abstract

Purpose

Statistics problem solving is complex and can be challenging for university students (Muis & Duffy, 2013). Because of the complexity of solving statistics problems, students must engage in self-regulated learning (De Corte et al., 2000). Undoubtedly, statistics problem solving is a demanding cognitive task, but it also includes affective components. As students solve complex problems, the most frequent emotion they experience is confusion, which can hinder learning (Di Leo et al., 2019). As Järvenoja and Järvelä (2005) demonstrated, students often lack the self-regulatory or emotion regulation skills to effectively resolve impasses during learning. As such, emotion regulation (Gross, 2014) is critical during complex problem solving. Accordingly, the purpose of this study was to develop a cognitive-emotive strategy training intervention to help students regulate and resolve confusion during statistics problem solving to improve learning outcomes.

Theoretical Frameworks

Pekrun’s (2006) control-value theory of achievement emotions and Gross’s (2014) process model of emotion regulation were the theoretical frameworks that informed our intervention. For the intervention, we targeted increasing students’ perceptions of control during learning and cognitive reappraisal of confusion.

Methods and Materials

One hundred forty students from various university statistics courses volunteered to participate and were randomly assigned to a control or intervention group. Students completed a self-reported measure of emotions about statistics (e.g., surprise, confusion, curiosity; Pekrun et al., 2017), a confusion regulation questionnaire (CRQ; Authors, 2022; competence development, reappraisal, suppression, other), and their perceptions of control about statistics problem solving (Muis et al., 2015) at pretest. Control group students then read a short text on the binomial distribution and solved increasingly complex binomial problems. For students in the intervention group, prior to being presented the short text and statistics problems, students were first presented a 15-min video that taught/modeled self-regulatory and emotion regulation skills, and then normalized confusion during learning. For all students, self-regulatory and emotion regulation processes were captured using a think/emote aloud protocol. After completing the problems, all students then completed the same pretest measures (emotions, CRQ, perceptions of control).

Results and Significance

Cronbach’s 𝛼 ranged from .78 to .92 for all variables. Results from path analyses revealed that the higher students’ perception of control during learning, the lower their level of confusion (β = -.52). Lower levels of confusion positively predicted performance on the statistics problems (β = .48) (𝜒2= 1.26, df = 1, p = .26, CFI = .99, RMSEA = .02). ANOVA results revealed that students in the intervention group reported the same level of confusion during problem solving, but higher perceptions of control over learning, used more cognitive reappraisal strategies, and achieved higher scores on the statistics problems compared to students in the control group.

Results provide further support for Pekrun’s (2006) control-value theory with regard to relations between appraisals and emotions wherein higher perceptions of control result in lower levels of confusion. These results also speak to the need to develop interventions to help students increase perceptions of control during complex learning tasks, and how to regulate confusion when it arises.

Authors

• Martina Calçada Kohatsu, McGill University

• Krista R. Muis, McGill University

• Reinhard Pekrun, University of Essex

• Gale Sinatra, University of Southern California

• Panayiota Kendeou, University of Minnesota

• Kristy A. Robinson, McGill University

• Alana Aiko Uilani Kennedy, Northern Arizona University