Faculty Advisor

Janice Gobert

Faculty Advisor

Ryan Baker

Faculty Advisor

Ivon Arroyo

Faculty Advisor

Michael Sao Pedro




Developing explanations is a key inquiry practice in national science standards (NGSS Lead States, 2013) and essential for learning science content (McNeill & Krajcik, 2011) and is conceptualized as consisting of three aspects: claims, evidence, and reasoning (Toulmin, 1958). However, students often have difficulty with these tasks (McNeill & Krajcik, 2011; Schunn & Anderson, 1999). Prior work by our group (Sao Pedro et al., 2014) has shown that auto-scaffolding in Inq-ITS (Inquiry Intelligent Tutoring System; Gobert et al., 2013) can help students acquire inquiry skills and transfer them to a new science topic. These data provide a rationale for the work presented, namely, designing, developing, and evaluating a real-time scaffolding approach for the development of the inquiry practices specifically for data interpretation and warranting claims, which, to us, underlie the explanation practices necessary for communicating science findings. Unpacking these practices can help us better understand, assess, and, in turn, scaffold them. Specifically, this work addresses the: (1) design of scaffolds for data interpretation practices; (2) efficacy of scaffolds for supporting these practices using a modified Bayesian Knowledge Tracing framework that captures the complexities of science inquiry, and (3) transfer of these practices within one science topic to another. Results from this work show that the developed scaffolds were effective in aiding students’ acquisition and transfer of the assessed practices. As such, this research builds on prior work on the nature of explanation (McNeill & Krajcik, 2011) as well as prior work on the assessment and scaffolding of science inquiry skills (Gobert et al, 2013; Sao Pedro et al., 2014).


Worcester Polytechnic Institute

Degree Name



Learning Sciences and Technologies

Project Type


Date Accepted



Restricted-WPI community only


scientific inquiry, data interpretation, scaffolding

Available for download on Monday, April 26, 2021