Causal Modeling to Understand the Relationship between Student Attitudes, Affect and Outcomes

We are using causal modeling to analyze relationships between pedagogical intervention, students’ attitudes, affective states, perceptions and outcomes, based on the data from a math tutor, Wayang Outpost. The causal model generated gives interpretable multi level interrelationships within the data variables identifying direct and indirect effects among them. We observed that among the four affective variables, confidence and frustration are more tightly linked with their performance and ability whereas interest and excitement are more related to their attitude and appreciation of math and tutor.