Control and value appraisals shape students' achievement emotions: a longitudinal analysis of multiplicative effects

Year: 2013

Author: Becker-Kurz, Betty, Pekrun, Reinhard, Frenzel, Anne C., Marsh, Herbert W.

Type of paper: Abstract refereed

Students' achievement emotions are critically important for their learning and achievement, suggesting that researchers should examine their antecedents. As addressed by Pekrun's (2006) control-value theory of achievement emotions, cognitive appraisals of control over and the value of achievement are thought to be primary determinants of these emotions. Specifically, the theory posits that perceived control and value combine in multiplicative ways in the arousal of achievement emotions. In the present research, we tested this proposition for students' enjoyment, pride, anxiety, anger, and shame.Using a longitudinal design, we investigated students' appraisals and emotions at two time points (beginning and end of school term). The sample consisted of 1,516 students (57.7% female) from 65 classes in 10 German secondary schools (grades 5-11). Appraisals and emotions were measured in the context of one of the students' core school subjects (German, English, mathematics and science). Measures included the short German version of the Self-Description Questionnaire (SDQ; Marsh, 1990, Marsh & O'Neill, 1984), Jacob's (1996) Achievement Value Scale, and the subject-specific version of the Achievement Emotions Questionnaire (AEQ; Pekrun, Goetz, Frenzel, Barchfeld, & Perry, 2011). Three sets of structural equation models with latent interactions were performed to examine main and interactive effects of control and value on students' emotions (within Time 1, within Time 2, and across Time 1/2). The results indicate that control and value have multiplicative effects on enjoyment, anxiety, anger, and shame within and across both time points. Within and across both time points, the AIC and BIC values were lower for structural equation models with latent interactions (t1/t2/t1-t2: AIC: 110,707.232 / 107,708.092 / 109,198.143, BIC: 111,156.800 / 108,151.545 / 109,657.279) than those for structural equation models not including latent interactions (t1/t2/t1-t2: AIC: 110,732.979/ 107,746.542/ 109,220.387, BIC: 111,172.375/ 108,179.962/ 109,669.135), which indicates that the structural equation models with latent interactions fit the data better (Kelava et al., 2011). These results support our study hypotheses and provide evidence that students' perceived control and value have multiplicative effects on their emotions, above and beyond main effects of these appraisals. From an educational perspective, the findings imply that interventions aiming to increase students' positive emotions and decrease their negative emotions should take both their control appraisals and their value appraisals into account.