Text

theses

When scientists develop knowledge about the world, they engage in a variety of complex epistemic processes (Allchin, 2011; Hardwig, 1985). They evaluate scientific models and evidence (Giere, 2004) and evaluate not only their own claims but the claims of others (Chinn, Rinehart, & Buckland, 2014) through the use of argumentation (Thagard, 2000). School science often omits the authentic epistemic practices of scientists, producing a false characterization of their work (Allchin, 2004; Chinn & Malhotra, 2002, Duschl, 1988). Science classrooms tend to be epistemically sterile environments (Goldberg, 2013) focused on unproblematic accounts of science (Allchin, 2004; Duschl, 1990). Recent calls for reform argue that there is a need for learning environment designs where students grapple with opposing perspectives and uncertainty like that found in the world outside of school (Britt, Richter, & Router, 2014). This research addresses these concerns in three parts. Chapter 2 presents a design case discussing four key design principles for engaging students with models and evidence in environments that embrace uncertainty and multiple, sometimes conflicting, perspectives. These decisions involve: identifying phenomena for students to investigate, designing for student engagement with modeling, developing evidence for use during modeling, and fostering productive disciplinary engagement (Engle & Conant, 2002). Chapter 3 examines how students use, evaluate, and re-evaluate evidence over time and how their ideas about one piece of evidence impact their ideas about other evidence. I present the results of a three-day model-based inquiry lesson with 7th grade students who investigated the possibility that some humans might be genetically resistant to HIV. Existing frameworks for evaluating student reasoning do not include evidence re-evaluation or the combination of pieces of evidence to construct a new body of evidence. I argue that normative accounts of good reasoning in science classes could be improved by taking both of these practices into account. Chapter 4 presents the results of a three-day modeling activity in which 7th grade life science students developed models of inheritance in response to multiple evidence sets. Students developed models that: were consistent with evidence, were internally consistent, increased in their use of causal mechanisms, and increased in their consistency with normative explanations of inheritance. Students’ abilities to correctly make predictions about novel inheritance problems significantly increased over time.

ETD_8327

Ph.D.

Includes bibliographical references

by Ronald W. Rinehart

doi:10.7282/T35Q508C

ETD doctoral

The author owns the copyright to this work.