Control-value theory and enjoyment of science: A cross-national investigation with 15-year-olds using PISA 2006 data

Highlights

• Control and value appraisals had large effects on enjoyment of science.

• The control-value interactions were statistically significant but very small.

• The relative importance of control and value on enjoyment varied among countries.

• The links between control, value and enjoyment were sensitive to gender and country.

Abstract

Control-value theory (CVT) explains how achievement emotions arise. CVT assumes control and value appraisals and their interaction are proximal antecedents of achievement emotions, and the structure of relationships between these constructs are independent of culture and gender. Using the Programme for International Student Assessment (PISA) 2006 data collected from nationally representative samples of 15-year-olds from 57 countries, this study attempted to test these assumptions for enjoyment of science by employing structural equation models with latent interactions. The results showed that control (science self-concept), personal value of science, and utility value (instrumental motivation to learn science) had large positive effects on enjoyment of science. The interactions were enhancing and statistically significant but effect sizes were very small. There were differences among countries with respect to the relative importance of control and value appraisals on enjoyment. The results showed that structural relationships can be considered equivalent across 57 countries and both genders.

Introduction

Enjoyment of science learning as an achievement emotion has a central role in predicting and mediating different constructs relevant to learning. Enjoyment of science learning is shown to predict engagement in science learning tasks (Hampden-Thompson & Bennett, 2013), self-control (Galla, Amemiya, & Wang, 2018), participation in extracurricular science activities (Lin, Lawrenz, Lin, & Hong, 2012; Nagengast et al., 2011), interest in science (Ainley and Ainley, 2011a, Ainley and Ainley, 2011b; Ainley & Hidi, 2014)⁠, science-related career aspirations and academic achievement (Jeffries, Curtis, & Conner, 2020), and collaborative problem solving (Camacho-Morles, Slemp, Oades, Morrish, & Scoular, 2019). Hence, an understanding of how enjoyment of science learning as an achievement emotion originates may help to deepen our understanding of students' learning experiences and processes. Moreover, a knowledge of the nature of the antecedents of enjoyment of science is highly relevant to the planning of interventions that foster enjoyment and promote learning.

Control-value theory (CVT) of achievement emotions (Pekrun, 2006b; Pekrun, Frenzel, Goetz, & Perry, 2007) is an educational psychology theory that explains how achievement emotions arise. CVT posits that achievement emotions occur as a consequence of control appraisals, value appraisals, and the interaction of these appraisals. Although CVT expects interactions between control and value appraisals, these interactions have been included in few studies. Moreover, most of the research had been conducted in Western settings (Pekrun, 2018). Furthermore, while CVT assumes that the structure of control and value appraisals would be universal (independent of gender and culture) (Pekrun & Stephens, 2010), very few studies have explored such generalizations. The purpose of this study is to address these limitations in the existing research by examining how control and value appraisals and their interaction predict enjoyment, one of the most intense and frequent achievement emotions experienced by students. This study also tests whether this prediction can be generalized in terms of gender and across 57 countries, in the domain of science learning for 15-year-old students, using secondary analyses of data from the Programme for International Student Assessment (PISA) 2006 (Organization for Economic Co-operation and Development, OECD, 2006).

Achievement emotions are directly related to achievement activities or outcomes (Pekrun, 2006b). Achievement refers to the quality of activities or their outcomes, referred to as object focus dimension, and can be evaluated with respect to specified standards of excellence (Paoloni, 2014). In addition to object focus dimension (activity versus outcome), CVT organizes achievement emotions into their valence dimension (positive versus negative) and activation dimension (activating versus deactivating). Furthermore, outcome emotions can be prospective or retrospective. For example, an achievement activity might involve conducting an experiment in a science class and the associated activity-based achievement emotion experienced can involve enjoyment of the science class. An achievement outcome can be the result of a science exam. The associated prospective outcome emotion can involve anticipatory joy if the student expects to obtain a good grade before the exam. The retrospective outcome emotion associated with the science exam can be retrospective joy about having experienced success in the exam after the student learns that she has received a good grade in the exam. In CVT, enjoyment is represented as an activity-focused, positive, activating emotion and involves experiencing pleasure in an activity. Enjoyment is distinct from the prospective emotion of anticipatory joy of expecting success or from the retrospective emotion of joy about having experienced success.

Achievement emotions are regarded as either state or trait-like constructs depending on the context (Bieg, Goetz, & Hubbard, 2013). For example, enjoyment during a science activity can be taken as a state like construct, whereas enjoying science lessons in general, can be regarded as a trait-like construct.

CVT posits that an individual's cognitive appraisals of control and value are the most proximal determinants of achievement emotions. When students regulate their control and value appraisals they can alter the achievement emotions they experience (Buff, 2014). Control appraisals pertain to the degree that an achievement activity and its outcomes are perceived to be controllable, and are usually indicated by competence perceptions such as self-concept of ability.

Value appraisals can be intrinsic or extrinsic (Pekrun, 2006b). Intrinsic value refers to evaluating an activity or outcome as interesting and important in its own right. Extrinsic values can come in different forms including achievement value, the personal importance of getting good grades and utility value, evaluating an activity or outcome as instrumental in acquiring a desired outcome (Putwain et al., 2018). Moreover, extrinsic values may involve the relevance of a domain to a person's everyday life (Goetz et al., 2014). In PISA 2006, “the value of scientific knowledge and inquiry for the students' current and future everyday life” (Pekrun, 2006a, p.21) is conceptualized as the personal value of science. Another subtle issue is the distinction between the value of a domain in general versus the value of studying the domain in school (Tytler, 2014). In PISA 2006, utility value items are worded with respect to studying school science, whereas personal value of science items are worded with respect to science in general. Previous research showed that control appraisals usually have large effects and intrinsic and utility value appraisals have moderate to large effects on enjoyment (Goetz, Frenzel, Stoeger, & Hall, 2010; Putwain, Pekrun, et al., 2018).

According to CVT, as a social cognitive theory, more distal factors (environmental antecedents) are significant. These factors include the characteristics and the cognitive requirements of the task, the degree of cognitive and emotional support the teacher provides, the overall learning atmosphere, and the broader social and cultural context within the school. These factors are posited to affect achievement emotions, predominantly through their influence on control and value appraisals. Achievement emotions in turn influence motivation, engagement, learning, and performance outcomes. Therefore, these relationships are reciprocal; environmental antecedents, cognitive appraisals, emotions, and their consequences all influence each other (Pekrun et al., 2007). For example, not only a student's self-concept of ability influences his or her achievement emotions, but also his or her emotions will shape his or her subsequent self-concept of ability beliefs.

CVT acknowledges the context-dependent nature of achievement emotions and posits that they are organized in domain specific ways. Contemporary approaches to emotions conceptualize them in relation to specific contexts (Griffiths & Scarantino, 2009; Gross, 2014) and control and value appraisals such as perceptions of one's skill, achievement expectations, and interest have been shown to be domain specific (Bong, 2001; Marsh, 1986). Therefore, emotions that are linked to control and value appraisals are also considered to be domain specific (Brunner et al., 2010; Goetz et al., 2014; Goetz, Frenzel, Pekrun, & Hall, 2006; Gogol, Brunner, Preckel, Goetz, & Martin, 2016). For instance, one can have different control and value appraisals in mathematics or in science.

According to CVT, control and value appraisals interact when producing an achievement emotion (Pekrun et al., 2007). In other words, the relationship between achievement emotions and control and value appraisals are both additive and multiplicative (Pekrun & Perry, 2014). Among the very few studies which explored this interaction, Goetz, Frenzel, et al. (2010) reported that for a sample of 50 undergraduates high scores of value amplified positive relationships between control and enjoyment. Similarly, using a sample of 579 fifth grade students, Putwain, Pekrun, et al. (2018) found neither an interaction effect between control and intrinsic value nor between control and utility value for enjoyment but detected a significant interaction for control and achievement value.

In CVT, the general structural features and causal mechanisms of emotions are assumed to be universal; they are regarded as inherent properties of the human mind (Pekrun, 2006b, Pekrun, 2009). Emotions depend on control and value appraisals independent of gender or culture. On the other hand, the content of the emotions, their frequency, and intensity may vary between genders or cultures. Similarly, depending on the extent that perceived control and values differ between genders or cultures, the resulting emotional experiences can differ as well. Hence, CVT again acknowledges context dependency and predicts that in different genders and cultural contexts, the structure of the relationships between control appraisals, value appraisals, and emotions are expected to be equivalent, whereas the intensity of each may be different.

Regarding gender, previous research found differences in the mean levels of enjoyment of mathematics favoring boys as well as perceived control and value (Frenzel, Pekrun, & Goetz, 2007; Pekrun, Frenzel, Goetz, & Perry, 2007; Goetz, Frenzel, Hall, & Pekrun, 2008; Holm, Hannula, & Björn, 2017). Pekrun, Goetz, Frenzel, Barchfeld, and Perry (2011) found that for psychology classes females reported higher levels of class-related enjoyment. Frenzel, Pekrun et al. (2007) detected minor discrepancies between the genders in structural relationships between the enjoyment of mathematics and prior achievement, control appraisals, and value appraisals, whereas Goetz et al. (2008) reported that the structural relations between enjoyment of mathematics and prior achievement and control appraisals were equivalent across genders, but they have not measured value appraisals. Similarly, Raccanello, Hall, and Burro (2018) found that the structural relations between enjoyment and control and value appraisals were invariant across genders for mathematics and literacy domains.

With respect to cross-cultural comparisons, Frenzel, Thrash, Pekrun, and Goetz (2007) found that Chinese students reported a greater enjoyment of mathematics than German students. However, the researchers did not test for structural equivalence of enjoyment and control and value appraisals across China and Germany because they did not measure control and value appraisals. Indeed, a thorough review of the literature did not locate any studies testing cross-cultural comparisons of the structural relations between control and value appraisals and enjoyment.

For many countries, by the age of 15 most students are approaching the end of their compulsory education (OECD, 2009). It makes students at this age an optimal population to study in respect of the cumulative effects of compulsory schooling on achievement emotions. By the age of 15 adolescents develop the cognitive abilities necessary to distinguish all types of achievement emotions and master the related appraisals of competence and value (Pekrun, 2017; Pekrun & Stephens, 2012). Compared to earlier ages, they can spontaneously report more emotions and their antecedents (Raccanello et al., 2018). They have spent many years in schools, receiving, interpreting, and incorporating feedback into their academic self-concept. Students who are successful in school incorporate their academic self-concept into their global self-concept by making academics part of their identity, whereas, students who fail in or see little utility in school reduce their level of identification with academics (Osborne & Jones, 2011). By the age of 15, the majority of adolescents have already decided whether they want to study science further (Venville, Rennie, Hanbury, & Longnecker, 2013). Most students at this age have reached the bottom of the decline in academic self-concept, enjoyment, and interest in academic subjects (Raccanello, Brondino, & De Bernardi, 2013).

The aim of the present study was both to examine the role of perceived control and value in the domain of science as proximal antecedents of enjoyment of learning science in 15-year-old students and to test the generalizability of this role across countries and genders. The research evidence on the cross-cultural generalizability of the links between control and value appraisals and achievement emotions is particularly scarce in the domain of science; hence, this study uses the most extensive database available to investigate this problem.

In this study, enjoyment was prioritized over the study of other achievement emotions for several reasons. First of all, the correlations between self-concept and enjoyment in comparison with self-concept and outcome based emotions such as pride are of a similar magnitude for science-related domains. For example, Goetz et al. (2014) reported that for physics the correlation between self-concept and enjoyment was 0.79 versus between self-concept and pride 0.82 for students in grade 11, which were similar to the correlations reported by Goetz, Cronjaeger, et al. (2010). Second, because enjoyment is an activity based emotion it is more readily related to learning tasks than pride, an outcome based emotion. Enjoyment of learning is related positively with flow experiences (total immersion in the task) and negatively with task irrelevant thinking (Pekrun, Goetz, Titz, & Perry, 2002). Enjoyment of learning can also help positively appraise the value of tasks and the person's beliefs of competence to solve them, leading to increased interest and motivation (Pekrun, 2017). Third, enjoyment is one of the most intense and frequent achievement emotions experienced by students (Putwain, Becker, Symes, & Pekrun, 2018) and the most often cited emotion by high school students (Raccanello et al., 2018). Finally, PISA 2006 only provides data on enjoyment but not pride.

Based on CVT the following research questions and associated hypotheses were investigated.

Research question 1: Do control and value appraisals and their enhancing interaction predict enjoyment of science?

Hypothesis 1

Perceived control and value appraisals of science and their synergistic (enhancing) interaction positively predict enjoyment of science. In the latent interaction structural equation model the path coefficients of control, value, and control value interactions will be positive.

Research question 2: How similar are the structural relations between control and value appraisals and enjoyment among different countries? Is the prediction in Hypothesis 1 generalizable over 57 countries?

Hypothesis 2

The relationship between perceived control, value appraisals, their enhancing interactions and enjoyment of science are structurally equivalent across countries.

Research question 3: How similar are the structural relations between control and value appraisals and enjoyment between genders? Is the prediction in Hypothesis 1 generalizable over genders?

Hypothesis 3a

The relationship between perceived control, value appraisals, their enhancing interactions and enjoyment of science are structurally equivalent across genders when the entire international sample is considered.

Hypothesis 3b

The relationship between perceived control, value appraisals, their enhancing interactions and enjoyment of science are structurally equivalent across 114 gender by country groups.