16:30 - 18:00
Parallel sessions 3
16:30 - 18:00
Room: HSZ - N4
Chair/s:
Luisa Schulz, Franziska Ingendahl
Research on metacognition investigates how people understand and regulate their own cognitive processes. This symposium addresses how metacognitive monitoring judgments are formed and how they influence effective learning. The first two talks focus on the underlying basis and accuracy of metacognitive judgments: Schulz, Bröder, and Undorf show that people integrate multiple cues when making metacognitive control decisions. Leipold and Berthold find that Judgments of Remembering and Knowing (JORKs) differ from traditional Judgments of Learning (JOLs) in memory processes, although the previously reported accuracy advantage of JORKs was not replicated. In the third talk, Schaper and Ingendahl present evidence on how metacognitive judgments shape item and source memory. The last two talks provide insights into more applied aspects of metacognition: Zawadzka and Hanczakowski show how feedback motivates learners to solve general knowledge facts themselves. Finally, Undorf, Ingendahl, Janson, Wissel, and Münzer demonstrate that JOLs predict learning behavior and success in a higher education learning setting. Together, the talks provide new insights into the mechanisms and consequences of metacognitive monitoring for learning and memory.
Submission 120
Judgments of Learning Predict Learning Success and Behavior in Higher Education
SymposiumTalk-05
Presented by: Monika Undorf
Monika Undorf 1, Franziska Ingendahl 1, Marc Philipp Janson 2, Samuel Wissel 3, Stefan Münzer 3
1 Technical University of Darmstadt, Germany
2 Karlsruhe University of Education, Germany
3 University of Mannheim, Germany
Numerous experimental studies have shown that judgments of learning (JOLs), people’s predictions of their future performance during learning, are moderately predictive of test performance and learning behavior. However, almost all previous studies have used relatively simple learning materials and low-stakes tests. Consequently, little is known about the predictive power of JOLs in ecologically valid learning environments. To address this gap, we collected JOLs in an intelligent tutoring system that university students used to prepare for their final exams in Statistics (Study 1, N = 90) and Educational Psychology (Study 2, N = 145) and retrieved objective indicators of learning success and behavior from the tutoring system. JOLs significantly predicted both exam performance and learning success within the system, even after controlling for GPA. Moreover, JOLs predicted how students allocated their study time. Together, these findings demonstrate that JOLs possess predictive validity in real-world learning contexts.