U-Shaped development of Newtonian concepts: Implications for pedagogical design and research practice

TL;DR

This study investigates the non-linear development of Newtonian concepts in students, revealing a U-shaped learning curve with periods of regression, and discusses implications for teaching strategies and research methodologies.

Contribution

It provides empirical evidence of a U-shaped developmental pattern in Newtonian understanding and offers a hypothesis explaining this phenomenon, influencing pedagogical and research approaches.

Findings

Evidence of non-monotonic conceptual development

Identification of regression periods in understanding

Implications for teaching and research practices

Abstract

In the years 2002-2004, the Learning by Design\texttrademark group at Georgia Tech conducted an investigation of the development of qualitative Newtonian concepts as a function of time. I describe the classroom context and method for this investigation, involving the coordination of multiple measures including diagnostic quizzes, structured interviews, ethnographic observations in class and performance assessments. This paper specifically concerns the results from the diagnostic quizzes, relying on other measures as supporting evidence. I present convincing evidence of conceptual gains from that data. I also show that when looked at in more detail, there is a clear non-monotonic conceptual development process in which there is a period of apparent regression of understanding following apparent satisfactory understanding. I consider prior work on this effect from developmental psychology, and propose a hypothesis to explain its presence in this data. I then consider how cognitive models converge on a similar explanation, and in the process contrast this interpretation with the currently extant interpretation of related work in terms of an interference effect. I close with consideration of implications of this effect for both pedagogical design and research practice.