From academic engagement to academic achievement: path analysis of freshmen achievement process.

De Clercq, Mikaël [UCL] Frenay, Mariane [UCL] Galand, Benoît [UCL]

Hitherto, many approaches of academic achievement among first year at university have been proposed in the literature and a host of predictors have been identified (Richardson et al., 2012). However, most studies evaluated the impact of these predictors one by one without taking their interrelations into account. Recent surveys shed the light on the insufficiency of such single path analyses to understand the complex processes of academic achievement (De Clercq et al., 2013; Pintrich, 2003) In this line, several authors underlined the necessity to consider background variables (past performance, parents’ education, gender…) and psychosocial factors together to fully grasp achievement issue (Allen et al. 2010, Dollinger et al., 2008, Robbins et al., 2004) and advocated that developing a more inclusive approach will lend deeper clarification of freshmen adaptation to the university. From a theoretical point of view, self-processes model could meet the needs for such an inclusive approach of academic achievement putting forward psychosocial factors while including background variables (Appleton et al., 2008). This model was initially framed to fit with secondary education (Skinner et al,. 1990). In this context, studies demonstrated the critical role of student engagement to predict achievement and dropout. Hence, this model essentially focused on engagement construct which is depicted as the most proximal predictor of achievement (Skinner, 2012). A tripartite conceptualization of engagement is endorsed including behavioral, cognitive and emotional dimensions (Fredricks et al., 2004). According to the model, engagement is directly determined by self-system processes -characterized by student’s motivational and social variables- and indirectly by background characteristics. The aim of our research is twofold. First, we investigated the fit of an inclusive modelling of achievement adapted from self-processes model. Second we undertook a more in depth investigation of the actual impact of engagement on academic achievement in higher education. Our sample consisted in 503 freshmen in psychology. Background characteristics were assessed through high school grade and socioeconomic status. Self-system processes were measured by academic self-efficacy beliefs, courses value and peer support. The three dimensions of engagement were respectively assessed by study time, deep strategy use and positive emotion and were measured twice, during the first (T1) and the second semester (T2). Finally, measures of achievement were taken two times through first and second semester grade. The procedure consisted in two stage process using path analysis. First, the fit of achievement process model was investigated. Second, using longitudinal design, the impact of engagement dimensions (T1&T2) on first and second semester grade was analyzed controlling for past performance. Preliminary results highlighted that inclusive model of academic achievement has good fit indices. However, initial self-system processes model is not the best fitting model and demand adaptation to fit to higher education context. Concerning the impact of engagement on achievement, both cognitive and behavioral engagements are consistently predictive of achievement even after controlling for past performance. This research substantiated the core impact of engagement on achievement in higher education and put forward an inclusive conceptualization of freshmen achievement process which could deal with current limitations in achievement literature.