Growth Curve Cognitive Diagnostic Models for Longitudinal Assessment

In recent years, cognitive diagnosis models (CDM) have been widely used to provide diagnostic feedback, including mastery/non-mastery of fine-grained skills. Most CDMs treat responses as if they were from a single time point and do not account for change in knowledge proficiency across time. Students’ skill knowledge, however, changes over time as they go through the process of learning; in this regard, it is important to understand their potentially diverse learning trajectories. By understanding students’ learning over time, educators can monitor students’ progress toward their respective learning goals and decide what to adjust to achieve better learning. Students themselves can also be informed on what skills they are specifically lacking and what they need to focus on next to reach their goals. In this talk, I propose longitudinal growth curve cognitive diagnosis models (GC-CDM) to incorporate learning over time into the cognitive assessment framework. The model is estimated using the marginal maximum likelihood (MML) method in Mplus. Relevant issues for estimation are discussed, e.g., the high-dimensional computation problem and model identification. Simulation studies show good parameter recovery. The model is also applied to real data using two datasets from multi-wave experiments designed to assess the effects of the Enhanced Anchored Instruction (EAI; Bottge et al., 2003) on mathematics achievement.

SeungYeon Lee is currently a postdoctoral innovation fellow at EdLab, Teachers College Columbia University. Her research interests include psychometrics, multilevel modeling and educational data mining. She is particularly interested in using quantitative methods to measure various aspects of students' learning. SeungYeon received her PhD in Quantitative Methods and Evaluation from UC Berkeley in 2017 and an MA in Statistics from Yonsei University, Korea in 2012.

Tuesday, February 20, 2018 - 2:00pm
PDF icon SeungYeon Lee presentation.pdf1.21 MB