Effect of self-regulated learning strategies on students’ achievement in science: A meta-analysis

Self-regulated learning (SRL) is not one single model or even a single theory. Almost anything can fit under the SRL umbrella because it is a multidimensional construct. Even scholars have not reached any consensus about the definition of self-regulated learning (Li et al., 2018). By definition, self-regulated learning refers to the self-directive process, knowledge acquisition, thoughts, feelings, behaviours, complicated method and belief (Zimmerman & Risemberg, 1997; Panadero, 2017). In the educational context, SRL is not only a process of students’ responsibility and understanding about their learning but also their ability to develop knowledge, skills, and attitudes (Boekaerts, 1999; Zimmerman, 2002). Since its introduction in the mid-1980s, self-regulation has been examined from several major theoretical perspectives including the behaviourist theory, phenomenological theory, information-processing theory, social cognitive theory, and social constructivist theory (Flavell, 1979; Farah, 2017).

There are several models of self-regulated learning such as social cognitive (Zimmerman, 1990, 1998), flexible learning (Boekaerts, 1992), four-stage SRL (Winne, 1996; Winne & Hadwin, 1998), process-oriented models of metacognition (Borkowski, 1996), and general framework SRL (Pintrich, 2000). Strategies in self-regulated learning vary according to the different SRL models proposed. The model by Pintrich et al. (1991) contains four categories of SRL: 1) cognitive (rehearsal, elaboration, organisation, critical thinking); 2) metacognitive; 3) resource management (time, study environment, effort regulation, peer learning, help-seeking); and 4) motivation (intrinsic motivation, extrinsic motivation, self-efficacy, and test anxiety). Meanwhile, Zimmerman’s model (2000) has three categories of SRL: 1) metacognitive (goal setting, planning, organising and transforming, seeking information, rehearsal and memorising); 2) motivation (self-evaluation, self-consequences; and 3) behaviour (environmental structuring, keeping records and monitoring, reviewing texts/notes/exams, and seeking assistance from peers/instructors/other resources). Most scholars agree with these 14 SRL strategies, namely (1) self-evaluation, (2) organisation, (3) transformation, (4) goal setting, (5) planning, (6) information seeking, (7) record-keeping, (8) self-monitoring, (9) environmental structuring, (10) giving self-consequences, (11) rehearsing, (12) memorising, (13) seeking social assistance, and (14) reviewing (Nodoushan, 2012). In this meta-analysis study, we conceptualised SRL as a strategy to enhance student achievement in science subjects.

Meta-analysis study on self-regulated learning in relation to student’s achievement in previous studies have focused on the function or impact of self-learning and the efficacy of teaching methods on learning language subjects and mathematics. Haller et al. (1988), for instance, reported the effect of learning strategies on achievement that had a relatively high effect size (es=0.71) with a standard deviation of 0.81. In the study, they reviewed 20 articles from the year 1975 to 1987, focusing on the effects of metacognitive teaching by reading quotes for language subjects. The findings revealed that metacognitive teaching was significant. Hattie (1996) studied the relationship between self-management, motivation, and self-concept with students’ achievements and found a moderate effect size (es=+0.57). The timeframe for the data sources was between 1968 and 1992. Chiu (1998) analysed 43 metacognitive interventions on the extraction of quotes by reviewing articles from 1979 to 1995 where an average effect size of 0.40 was obtained.

The results of this study revealed that metacognitive interventions had a moderate effect on students’ achievement. Dignath et al. (2008) in their meta-analysis which was performed on 74 studies from 1992 to 2006 concerning self-study intervention among primary and secondary school students also reported an average effect size (es=+0.61). They also reported significant differences in the effect size for Mathematics, reading and writing among primary and secondary students. Meanwhile, Abrami et al. (2008) reported the effect size value of (es=+0.341) for 117 studies which involved 20,698 study samples with 161 computational size calculations in their study that examined the effect of the teaching method in promoting critical thinking skills based on the measurement of student self-regulated learning. A systematic literature review by Wang and Sperling (2020) also demonstrated that SRL intervention positively affected school-age learning in Mathematics.

Meanwhile, no empirical finding for the effect size analysis was reported in the SRL studies on achievement by Zimmerman and Schunk (2001) and SRL and science education by Schraw, Crippen and Hartley, (2006) and Zohar and Dori (2011). A recent meta-analysis study by Bembenutty and DiBenedetto (2013) on three SRL strategies (academic delay of gratification, homework self-regulation, and help-seeking) showed low to medium effect size on science achievement. Previous studies have proven the effectiveness of the role of self-regulated learning in science subjects (Cheng & Tsai 2020; Li et al., 2018; Olakanmi & Gumbo, 2017). SRL has been found not only to enhance the mastery and understanding of the concepts in science/chemistry and assist students to answer problem-solving questions, improve science-motivated learning as well as obtain better scores in tests, it also promotes a better sense of responsibility towards higher learning.

Levinson (1981) stated two essential variables that contribute to learning outside and inside the classroom, namely teacher and students’ self-awareness. Students with a high level of awareness are more likely to succeed (Edomwonyi & Avaa, 2011; Kihwele, 2014; Agustiani, Cahyad & Musa, 2016; Olakanmi & Gumbo, 2017). Moreover, the efficacy of self-regulatory learning can be linked to students’ achievement. Academic achievement is important in indicating a student’s progress in comparison to others such as whether the student has reached a high level of excellence in the academic field. Academic achievement reveals the students’ capacity in understanding the teaching materials. Students’ achievement depends on the strategies and resources used in relation to their learning styles. However, it is impossible to achieve success without an effective learning strategy plan because the achievement of everyone is influenced by the process used to learn the subjects, including that of science subjects. In addition, there is a lack of comprehensive meta-analysis studies that focus on science subjects. Therefore, the current meta-analysis research intended to look for the best self-regulated learning strategies for student achievement in science subjects.

This meta-analysis review was aimed at addressing the following research questions: 1) Which learning strategy is the most effective in enhancing students’ achievement in science subjects? and 2) What are the extent and circumstances of the impact of the strategies on students’ achievement in science? The objectives of the study was threefold. First, we intended to summarise the evidence on the impact of self-regulatory learning strategies on students’ achievement in science. Second, we sought to examine how effect sizes differ in relation to the impact of self-regulatory learning strategies on students’ achievement in science. Third, we wanted to know how the different types of moderator variables in the self-regulated learning strategies (types of science subject, categories of SRL) influence students’ achievement in science.