What Are We Talking About? The Jingle-Jangle Fallacy in Motivation/Interest Theories and Measurement

• In Event: Interest Development, Expectancies, and Values: Insights, Challenges, and Prospects

Abstract

There has been a proliferation of theories and related constructs/measures in the social-cognitive perspectives on motivation and interest development, leading to a jingle jangle fallacy problem across measures and theoretical frameworks. Our own work is grounded in the expectancy X value grand theory of motivated behavior (Eccles et al.,1993; Eccles & Wigfield, 2002) . We have identified one set of key social cognitive influences on people’s selection of, and then engagement in, skill-based, achievement-related tasks and activities focused on two fundamental questions (‘Can I do this?’ and ‘Do I want to do this?’). Other scholars have often generated quite similar construct lists and associated measures (Wigfield & Cambria, 2010). We believe it is time to integrate across theoretical perspectives and measured constructs in order to identify the key latent constructs. Given this, the field should be able to move forward by using more person-centered and complex systems data-analytic techniques to identify how these constructs interact to motivate and sustain both interest and skill development. We will draw upon our recent factor and person-centered analytic work to both illustrate the problem and model some potential solutions.

In study 1, we longitudinally assessed a set of motivational constructs drawn from Expectancy Value Theory, Achievement Goal Theory, and Self-Determination Theory in an online course of approximately 100 undergraduate students at UCI (Lee et al, unpublished manuscript). Using both EFA and CFA on both Time 1 and Time 2 assessments (separated by several weeks), we identified substantial overlap in the measures of subjective task value drawn from these various theories – overlap that replicates over time – supporting our concern over the jingle jangle fallacy in our field. In a second study based on our Childhood and Beyond longitudinal study of sustained engagement in specific skill-based activities (Wan & Eccles, unpublished manuscript), we found that well developed interests may be best assessed using a latent construct based simultaneously on several different measures of ability self-concepts, subjective task values and time currently being spent on the task, supporting the idea that well developed interest is an emergent property of multiple beliefs and behaviors. Clearly before the field can work on the ways in which these latent constructs interact with each other at the individual level to influence behavior both at one point in time and over time, more rigorous psychometric work is needed to identify the similarities and differences across theories of motivation and interest.

Finally, drawing upon our own ongoing longitudinal studies (gathered over the last 40 years), we will show how even within Expectancy Value Theory, there are reliable and replicable person-level differences in the ways in which expectancy and subjective task value constructs aggregate to predict well developed interests, college majors and career choices over time. We will discuss how these findings illustrate the need for more person-centered and more complex type of analyses to move this area forward in productive ways.

Authors

• Jacquelynne S. Eccles, University of California - Irvine

• Hye Rin Lee, University of California - Irvine

• Peter McPartlan, San Diego State University

• Osman Umarji, University of California - Irvine

• Sirui Wan, University of California - Irvine

• Allan L. Wigfield, University of Maryland - College Park