Learners' mental models of chemical bonding.

Abstract

The research reported in this thesis comprised a cross-age inquiry of learners' mental models for chemical bonding. Learners were chosen purposefully from three academic levels-senior secondary school (Year-13, age range 17-18 years old), undergraduate (age range 19-21 years), and postgraduate (comprising MSc and PhD; age range 22- 27 years). The principal research goal was to establish learners' preferred mental models for the concept of chemical bonding. Other research goals were to establish if and how learners made use of analogy to understand chemical bonding and to establish the prevalence of learners' alternative conceptions for chemical bonding. The research inquiry was conducted from within a constructivist paradigm; specifically the researcher ascribed to a social and contextual constructivist belief system.Based on a review of the science education literature a decision was made to classify mental models into four classes according to the typology of Norman (1983), namely, the target system, a conceptual model, the users' or learners' mental model and the scientists' conceptualisation. A conceptual theme for the inquiry was developed based on this typology resulting in the identification of target systems-metallic, ionic and covalent bonding. Subsequently, target models for each of the three target systems were identified, namely, the sea of electrons model and the band theory for metallic bonding; the electrostatic model, and the theoretical electrostatic model for ionic bonding; and the octet rule, the valence bond approach, the molecular orbital theory and the ligand field theory for covalent bonding. A conceptual model, consisting of a summary of the salient points of the target models, was developed by the researcher. Once validated by four of the instructors involved in the inquiry, this formed the scientists' conceptualisation for the target models.Learners' mental models were elicited by the use of a three phase semi-structured interview protocol for each of the three target systems based on the translation interface developed by Johnson and Gott (1996). The protocol consisted of showing participants samples of common substances and asking them to describe the bonding in these materials. In addition, participants were shown Interviews About Events (IAE), focus cards which depicted events involving chemical bonding or contained depicted models of bonding for the three target systems. Transcriptions of audio-tapes combined with diagrams produced by the participants formed the data corpus for the inquiry. Learners' mental models were compiled into inventories for each of the target systems. Examination of inventories enabled identification of commonality of views which were validated by four instructors-two instructors from the teaching institutions involved in the inquiry, and two instructors independent of the inquiry.The research reported in this thesis revealed that learners across all three academic levels preferred simple or realist mental models for chemical bonding, such as the sea of electrons model and the octet rule. Learners frequently used concepts from other more sophisticated models to aid their explanations when their preferred mental models were found to be inadequate. Senior level learners were more critical of mental models, particularly depicted models provided on IAE focus cards. Furthermore, senior level learners were able to describe their mental models in greater detail than their younger counterparts. However, the inquiry found considerable commonality across all three levels of learner, suggesting mental models are relatively stable.Learners' use of analogy was classified according to Dagher's (1995a) typology, namely, simple, narrative, peripheral and compound. Learners' use of analogy for the understanding of chemical bonding was found to be idiosyncratic. When they struggled to explain aspects of their mental models for chemical bonding, learners made extensive use of simple analogy, that typically involved the mapping of a single attribute between the target and source domains. There did not appear to be any correlation between academic ability or academic level and use of analogy. However, learners made greater use of compound analogy for the target systems of metallic and ionic bonding, mostly as a result of the use of analogical models during instruction.This inquiry revealed prevalent alternative conceptions for chemical bonding across all three academic levels of learner. This is a somewhat surprising result considering that the mental models preferred by learners were typically simple, realist models they had encountered during instruction. Learners' alternative conceptions often concerned simple conceptions such as ionic size, the presence of charged species in non- polar molecular compounds, and misunderstandings about the strength of bonding in metals and ionic substances. The inquiry also revealed widespread confusion about intermolecular and intramolecular bonding, and the nature of lattices structures for ionic and metallic substances.The inquiry resulted in a number of recommendations. It is proposed that it may be more beneficial to teach less content at the introductory level, that is, delivering a curriculum that is more appropriate for non-specialist chemistry majors. Hence, one recommendation is for instructors to examine the intended curriculum carefully and be more critical regarding the value of inclusion of some course content. A second recommendation is that sophisticated models of chemical bonding are better taught only at advanced stages of the degree program, and that teaching from a contructivist view of learning may be beneficial. The third recommendation relates to the fact that learners spontaneously generated analogies to aid their explanations and conceptual understanding, consequently, learners may benefit from greater use of analogy during instruction.