Influenced by Toulmin's idea of conceptual change, Posner et al. [#!posner-82!#] propose a theory of conceptual change for science education. They base their analysis on a study involving students' learning of special relativity via traditional instruction. Posner et al. suggest that a change in existing conception requires students to be faced with anomalies that create a dis-satisfaction in the existing conception and thereby make space for a new conception. Students also need to find the new conception intelligible, plausible and fruitful. Hewson, cited in [#!confrey!#], elaborates that when competing conceptions are presented children raise or lower the status of one conception relative to another. Posner and Strike later modified their model to reduce its ``excessive rationality'' by allowing for contextual influences.[#!cobern!#]
Dykstra et al. [#!dykstra!#] propose a taxonomy of conceptual change, based on Posner's general strategy. They describe three types of conceptual change viz. differentiation, class extension and re-conceptualization. Differentiation refers to the emergence of new conceptions from existing conceptions, as for example velocity and acceleration from ideas of motion in kinematics instruction. Class extension refers to those cases where the existing concepts considered different are subsuming concepts. For example, rest and constant velocity are equivalent in the Newtonian point of view. Re-conceptualization refers to those cases where a significant change in the nature of and relationship between concepts occurs. For instance, changing from force implying motion to force implying acceleration.
Chi et al. [#!chi-92!#] have proposed an alternative theory of conceptual change based on ontological categories. They base their theory on the supposition that the entities of the world can be classified into three primary ontological categories viz. `matter', `processes' and `mental states'. `Processes' are further categorized into `events', `procedures' and `acausal processes'; matter into `natural kinds' and `artifacts'. These are distinct in that the attributes of one category cannot be applied to those of another. Chi et al. suggest that misconceptions arise when a concept is assigned an incorrect ontological status. For instance, heat, force, light etc. are assigned the status of `matter'. According to them, conceptual change occurs when a concept is reassigned to an ontologically distinct category.
Many researchers in this tradition have attempted various methods of facilitating conceptual change. Major among these are the conceptual-conflict method and the anchor-bridges method. In conceptual conflict, the student is posed with a counter-intuitive situations which impels a need to re-think the previously held conception.[#!stavy!#][#!nussbaum!#] The anchor-bridges model on the other hand starts with a situation where most children intuitively reason correctly. It then proceeds across several conceptual bridges to a target situation where previously most students have exhibited misconceptions [#!minstrel!#] [#!clement-89!#] [#!thijs-95!#]. For example, to teach forces on bodies at rest, a series of experiments have been suggested [#!thijs-95!#]. In the first series students have to apply either pulling or pushing forces on a cart that is forced to remain at rest by hand or by a hanging mass. The next two series of experiments have a bridging character. Students start with the question of whether the rest-rule applies also when the cart is stopped by a wall and a block is kept at rest by friction. In the bridging series the existence of a balancing force is made plausible by making its effect visible. The bending of a flexible ruler and of the hairs of a brush aim at triggering the students' discovery of the normal force and the force of friction respectively.