Really Not That Different in Theory? A Meta-Analytic Control-Value Approach to Emotions in Technology-Based Learning

• In Event: Emotions in Technology-Rich Learning Environments: Theory, Measurement, and the Future

Abstract

The control-value theory of achievement emotions (CVT; Pekrun, 2006) provides an integrative framework for the study of emotions across different learning settings. The CVT accounts for modality-specific effects of technology-based versus non-technology-based learning settings (TBL) on the emotional mechanisms underlying learning by considering situation-specific appraisals as key triggers of these affective processes. As such, Daniels and Stupnisky’s (2012) conclusion that CVT is useful for explaining emotions in TBL seems valid. The aim of the current study is to solidify this claim through a synthesis of existing evidence regarding causes and effects of emotions in TBL using CVT as an overarching framework.

To this end, we adopted an inductive meta-analytic approach (Wilson, 2009) and conducted a comprehensive literature search to identify peer-reviewed studies addressing emotions and TBL in nonclinical samples. Sources included (a) PsycINFO and ERIC, (b) proceedings of the Intelligent Tutoring Systems, Artificial Intelligence in Education, and Educational Data Mining conferences, (c) manual journal searches (International Journal of Learning Technology, IEEE Transactions on Affective Computing), and (d) references of an edited book (Calvo & D’Mello, 2011) and chapter (Graesser, D’Mello, & Strain, 2014).

Employing emotion definitions provided by Pekrun, Goetz, Frenzel, Barchfeld, and Perry (2011; Pekrun & Meier, 2011), two researchers conducted an initial screen of the 550 candidate articles collected in our database. This led to the selection of the following emotions as the focal constructs of this review: (a) anxiety, (b) enjoyment, (c) anger/ frustration, (d) boredom, and (e) confusion (227, 129, 91, 64, and 57 articles, respectively). The final coding scheme was derived from the remaining studies and included (a) basic study and sample descriptors, (b) type of learning technology (D’Mello, 2013) and learning context (e.g., subject domain; TBL vs. blended learning), (c) correlates of emotions central to CVT (e.g., perceived control and value, self-regulation, achievement; Figure 1), and (d) methods used to assess emotions and their correlates. Studies were double coded by two researches and effect sizes were extracted in the form of correlation coefficients. Calculation of mean effect sizes and moderator analyses were carried out within a random-effects framework (REM; in the final analyses, effect size dependency will be handled using REMs proposed by Hedges, Tipton, & Johnson, 2010).

First findings reveal that research on emotions in TBL has evolved and grown tremendously over the past decades (Figure 2), both in terms of the range of emotions investigated as well as underlying theoretical approaches (Table 1), and that this research can be meaningfully integrated using CVT. Furthermore, initial results corroborate CVT’s assumptions regarding relations of anxiety and enjoyment with achievement (Table 2) and mirror those reported for traditional learning environments (e.g., Pekrun, Goetz, Titz, & Perry, 2002; Seipp, 1991). The findings confirm Daniels’ and Stupnisky’s (2012) observation that CVT can be extended to the growing field of TBL and help inform the design of emotionally sound technology-based environments (Lester et al., 2014). In providing integrative evidence, our study builds a foundation for the design of such environments and identifies directions for future research.

Authors

• Kristina Loderer, University of Munich

• Reinhard Pekrun, University of Munich

• James Lester, North Carolina State University