Desktop version

Home arrow Philosophy

  • Increase font
  • Decrease font

<<   CONTENTS   >>

Person-centered Approaches to Explaining Students’ Cognitive Processing Strategies


This chapter will present what is, as yet, a very small niche within the strategic processing research literature: the (potential) role of person-centered analyses for strategic processing research. This chapter is organized into three sections. The first aims to situate person-centered quantitative research methodologies within the plethora of analytical approaches that are commonly pursued. Then the reader is introduced to key person-centered research that has been undertaken, establishing the current state of the field and how it might continue to develop. The final section of this chapter makes a case for person-centered analytical approaches as a viable means of relating and integrating some of our processing strategy theories into a more comprehensive picture of how individual differences and the environment interact across a learners’ knowledge, skills and motivation-beliefs development.

Having stated the goals, it is important to make clear what the chapter will not attempt to do. It will not be a “how to” guide for researchers who lack sufficient experience with classical statistics. A grounding in statistics, both in “pen and paper knowledge” and experience designing and analyzing research, is necessary for this chapter to be of any real use and is directed at such readers. For readers seeking to brush up on the basics, Howell (2016) is suggested. For readers interested in a firm (detailed) grounding in the analytical methods discussed in this chapter you are referred to Hagenaars and McCutcheon (2002).

As a start, it is important to address the person-centered methods which will be the focus of this chapter. It is therefore relevant to note that all of the research the author has undertaken, and much of the recent research discussed, has been done with Latent Profile Analysis (LPA). LPA is consistent with common factor analysis (FA), whereby the covariation of variables (observed) is explained by continuous variables (latent). Bauer and Curran (2004) sum up the difference between LPA and FA by stating that “the common factor model decomposes the covariances to highlight relationships among the variables, whereas the latent profile model decomposes the covariances to highlight relationships among individuals” (p. 6).

Clustering approaches (i.e., K-mean) are historically (Steinley, 2006), and still currently, more popular, and have a demonstrated propensity for recovering accurate classification patterns (Steinley & Brusco, 2011). However, latent profile analyses have a number of demonstrated strengths which have made it the authors preferred approach to person-centered analysis. Two key strengths of LPAs are allowing partial membership into multiple subgroups rather than a rigid assignment to one and the many longitudinal techniques that a latent profile analysis approach makes possible (for a detailed list of the advantages of latent profile analysis relative to standard clustering approaches, please see the supplement to Morin et al., 2017).

<<   CONTENTS   >>

Related topics