Desktop version

Home arrow Philosophy

  • Increase font
  • Decrease font

<<   CONTENTS   >>

Section IV: Analyzing Strategic Processing

Variable-centered Approaches

Individuals’ strategic processing, or strategy use, is a subject that has fostered increased research in recent years (Dinsmore, 2017). Effective strategic processing is beneficial in accomplishing tasks in a variety of domains. Students intentionally, purposefully, and effortfully use strategies to learn while navigating through content (Cho, Affer-bach, & Han, 2018). Strategic processing has been linked to achievement outcomes, though the type and level of strategic processing matters in context (Dinsmore, 2017). Strategic processing is a valuable skill for learners of different ages. For example, in early childhood, learners use strategies to acquire and remember information (Nida, 2015). As another example, children in elementary school demonstrated they were able to respond to feedback in order to enhance strategy use in a dynamic testing situation (Resing & Elliott, 2011). Strategic processing affects learning outcomes for middle school students (Greene & Azevedo, 2009) and high school students, as well (Parkinson & Dinsmore, 2018). Additionally, the results of a recent meta-analysis showed that strategy use is a key variable associated with achievement in higher education (Schneider & Preckel, 2017).

This recent increase in research on strategic processing has led to a growing awareness that cognitive processes are not constrained by developmental stage, but are dynamic and malleable (Dinsmore, 2017). Because strategic processing is malleable, this provides the opportunity for students to learn more beneficial strategies. Researchers studying dynamic strategy use have capitalized on the malleability of strategic processing to identify and encourage students to use more beneficial strategies (e.g., practice testing and distributed practice) that encourage quality learning and achievement outcomes, rather than less beneficial strategies that encourage more shallow processing of the information (e.g., highlighting and rereading; Deekens, Greene, & Lobczowski, 2018; Dunlosky, Rawson, Marsh, Nathan, & Willingham, 2013).

People’s strategic processing can change over time within and across learning tasks (Dinsmore, 2017). Given this, research methods that are capable of capturing dynamic processing are necessary (Dinsmore & Zoellner, 2018). However, such methods can produce a lot of data that can be challenging to analyze. Depending on the nature of the data and the research aims, these data may be analyzed in many ways, including variable-centered analyses (Laursen & Hoff, 2006) that allow for an understanding of relations among numerous variables of interest (e.g., strategic processing, motivation, learning). In this chapter, first we outline the differences between variable-centered and person-centered analyses in the context of studying strategy use and strategic processing. Then we discuss the different kinds of variable-centered analyses that can be used to understand and model strategy use and strategic processing data. Then, we conclude with a summary of observations about variable-centered analysis use in the extant literature, as well as practical implications and future directions.

Variable-centered versus person-centered approaches

There are two prevailing approaches to quantitative data analysis: variable- and person-centered. In statistical parlance, researchers employ variable-centered analytical methods to describe the associations between variables within populations that are assumed to be homogeneous regarding those variables’ ability to predict a dependent outcome (Laursen & Hoff, 2006). These approaches involve analyses of relations between variables to produce a summarization of these relations, within a given set of parameters, to describe the entire population (Howard & Hoffman, 2018). For example, an R2 statistic summarizes how much variance in an outcome variable can be explained by a set of predictor variables, for a particular sample assumed to represent a larger population. Variable-centered statistical methods include correlation, regression, path analysis, structural equation modeling, and growth models. Variable-centered methods can be used to investigate the associations between people’s anxiety, knowledge of content, and frequency of deep strategy use. The hypotheses for this investigation might be: “There will be a negative relationship between anxiety and frequency of deep strategy use” and “There will be a positive relationship between people’s knowledge of content and frequency of deep strategy use.” A variety of different kinds of research questions pertaining to strategy use can be studied, including research questions about the efficacy of interventions (Yoon & Jo, 2014), strategy use change over time (Carr, Taasoobshirazi, Stroud, & Royer, 2011), the relation between strategy use and performance or knowledge gains (Greene, Deekens, Copeland, & Yu, 2018), and research questions pertaining to the relationship between strategy use and other learning phenomena (e.g., motivation; Bernacki, Byrnes, & Cromley, 2012).

Alternatively, person-centered methods use the relations between observed variables along with differences between individuals to identify multiple homogeneous subpopulations within a larger heterogenous population (Fryer & Shum, this volume). Also inherent in these methods is the identification of the appropriate number of emergent subpopulations needed to optimize the accuracy of the resulting population summary. Common person-centered approaches include latent class, latent profile, and cluster analysis (Howard & Hoffman, 2018; Laursen & Hoff, 2006). Person-centered analyses might be used to answer research questions such as, “Are there two or more groups of participants in this sample that systematically differ in their use of five common studying strategies?” Person-centered analyses are often preferable to variablecentered analyses when the ratio of participants to strategic processing variables is small, or when researchers are interested in identifying homogeneous subgroups for further analysis. Once these groups have been identified, they can be compared across a number of other covariates or criterion variables, such as prior knowledge or academic performance. Both person- and variable-centered analyses are viable methods for understanding strategic processing. In this chapter we focus on variable-centered approaches.

<<   CONTENTS   >>

Related topics