Doctor of Philosophy (Ph.D.)

Conducted in the context of learning monohybrid inheritance (a biological phenomenon explained by multi-level concepts), this entire study focused on the examination of one design assumption (i.e., students’ prerequisite knowledge) and two design decisions (i.e., collaborative participation structure and teacher-led compare and contrast) of the Productive Failure (PF) learning design which consisted of a generation and exploration (G&E) phase, followed by a consolidation and knowledge assembly (CKA) phase.

A total of 399 ninth- and 10th–grade students from a secondary school in Singapore participated in this study. Two Biology teachers from the same school and I were also involved. The nonequivalent control group design was adopted and the entire research study adhered to the following general research procedures, that is, pretest, study implementation (with inclusion of self-reported mental effort and lesson engagement surveys), and posttest.

The outcomes from the first study on provision of prerequisite knowledge showed that students, with prior instruction on relevant micro-level concepts, were able to generate more diverse RSMs. However, the RSM diversity did not confer upon students any learning advantage, and one possible reason could be the activation of inappropriate conceptions during the G&E phase that were not adequately addressed by the teacher’s direct instruction.

Based on the second study on provision of prerequisite knowledge, it was found that being part of any participation structure (i.e., individual or collaborative engagement) enabled students to generate and explore diverse concepts and elaborations (at both macro- and micro-levels) to a similar extent. Despite the similar generation and exploration capacity, the students individually engaged in the G&E phase were better able to apply their acquired concepts to solving advanced problems. Such outcomes could have arisen from certain social (e.g., downward matching) and/or cognitive factors (e.g., idea fixation) found in the collaborative participation structure that might have upset students’ overall learning.

The final study demonstrated that students, engaged in teacher-led compare and contrast, were significantly better in the acquisition of the targeted concepts, which may be attributed to having a teacher who is better at discerning critical conceptual features, explaining and elaborating on these various conceptual features, and further help in organizing and assembling them into the targeted conceptions. Although teacher-led C&C was significantly better for conceptual acquisition, the study further ascertained if the design of the student-led C&C activity could also be enhanced for learning. It was found that the further inclusion of support provisions (e.g., collaborative participation structure or teacher feedback) for the student-led C&C activity might not be necessary, as long as there is prior exposure to key conceptual features.

The collective outcomes from all three studies suggested that the PF learning design, in the context of a multi-level conceptual scientific phenomenon, might need further design considerations (e.g. teacher-led lecture on targeted concepts during the CKA phase that emphasizes key conceptual connections). However, the generalizability of these further design considerations might be further constrained by certain limitations (e.g., small sample size) found in the study. Some future work that extends from this research may entail the examination of various support structures (possibly infused into the G&E phase) on students’ overall learning from PF.

While there is a growing body of evidence on the efficacy of having students solve problems on novel concepts prior to receiving explicit instruction, little is known about the efficacy of having students find problems in order to prepare them to learn from the subsequent explicit instruction of target science concepts. Premised on a review of literature from two major sources, namely Preparation for Future Learning and past studies in science education involving students asking questions during the learning process, this doctoral research comprises a series of four Pilot Studies and three experimental Main Studies on problem-finding among fifth-grade mixed ability students in two elementary schools. Conceived as an exploratory research on preparatory problem-finding, the Main Studies were carried out to examine three issues.

Main Studies 1, 2A and 2B examined the first and overarching issue on whether problem-finding, with and without producing provisional answers to the problems, helped prepare students to learn target science concepts from the subsequent explicit instruction. The second issue, addressed specifically by Main Study 1, revolved around the effect of manipulating the order of student problem-finding activity and explicit concept instruction on students’ learning of the target science concepts. The third issue was related to the question of how students’ participation in the problem-finding activity, that is individually or collaboratively, might affect students’ learning of the target science concepts.

Findings from the Main Studies yielded some preliminary evidence on the efficacy of problem-finding on students’ concept learning. Specifically, findings of Main Study 1 revealed that students who collaboratively found problems with provisional answers before and after explicit concept instruction improved on their target concept test performances whereas improvement of test performance for students who received only explicit instruction was not conclusive. Findings from Main Study 2A and 2B also suggested that getting students to produce provisional answers to their own problems did not appear to confer them an advantage in learning the target concepts compared to peers who only found problems without corresponding provisional answers. Findings also suggested two previously unexamined factors, namely duration of problem-finding vis-à-vis explicit target concept instruction, and contrasting cases quantity in the problem-finding task, might have moderated the efficacy of problem-finding.

Pulling together and synthesizing from findings of individual Main Studies led to the conception of fresh lines of inquiry which, with future studies, can potentially lead to two new theorizations. The first, preliminary termed co-dividual learning, challenges the traditional dichotomous view of designing for individual or collaborative learning. This notion may be concretized and rendered useful in the form of a set of cognitive process based criteria to help determine when, for an instructional activity, it might be better for learners to collaborate, and when learning individually might be more ideal. The second theorization centers on the coupling of problem-finding and problem-solving processes to constitute a preparatory activity that might be more efficacious than either problem-solving or problem finding alone in preparing students to learn from subsequent explicit concept instruction.

While it must be recognized that conclusions from the studies reported in this thesis are tentative and hence premature to attempt any generalization of the findings, this research as a whole has surfaced some assumptions underlying extant research on Preparation for Future Learning, as well as laid a clear path for future empirical work designed to strengthen the evidential base of the case for problem-finding in learning science concepts. It has also given rise to two lines of inquiry that when carefully pursued can potentially lead to new theorizations on a nondualist mode of student participation in learning activities, and on a coupling of two distinct problem-based processes for stronger preparatory activity effect.