Education-line Home Page

An explanatory model for conceptual development during A-level chemistry.

Keith S. Taber

Homerton College, Cambridge


Paper presented at the British Educational Research Association Annual Conference, University of Sussex at Brighton, September 2 -5 1999.

Roehampton/Surrey symposium

Teaching, learning and relevance in classroom science


Acknowledgement: The work described was undertaken by the author when teaching at Havering College of Further and Higher Education

Current address for correspondence:
Dr. Keith S. Taber
Senior Lecturer in Science Education
Homerton College, Cambridge

(c) Keith S. Taber 1999

Extract, from the results section:

In general A level students’ developing understandings of chemical bonding concern a transition from one conceptual framework - the ‘octet framework’ - that typically characterises student thinking at the start of the course - to another, an electrostatic framework. The latter framework reflects (at least, to a greater extent) orthodox scientific understanding, whereas the former is a common alternative conceptual framework.


This paper considers how conceptual development in science may be realistically conceptualised. Research based on sequences of in-depth interviews with volunteer colearners (A level chemistry students) provides insights into aspects of developing understanding of a key scientific topic (chemical bonding). This research supports a view that our models must explain both the presence of stable coherent conceptual frameworks in cognitive structure, and the tendency of students to apparently call upon a menu of explanatory schemes to make sense of the science curriculum. Conceptual development is here seen as a shift in the extent to which alternative frameworks are applied.


This paper presents an explanatory model which is intended to make sense of data concerning changes in student conceptions during A level chemistry. However, the specific findings discussed are considered in the light of the wider literature into conceptual change in learning science.

Conceptual development in science.

The present study was undertaken from a constructivist perspective, a viewpoint which is informed by the great abundance of research into ‘alternative conceptions’ in science. Whilst I consider this canon of work to be of great importance, the cataloguing of students’ ideas is only of practical importance when it is seen against a sound theoretical understanding of the significance of such ideas. That is, the status of student conceptions (major blocks on learning?, resources for learning?, irrelevancies to learning?) is tied to our beliefs about the nature of cognitive structure and the mechanisms of conceptual development. At present there is no generally accepted consensual view about either of these key topics.

The nature of cognitive structure.

In my work (Taber, 1997a) I have used the term ‘cognitive structure’ to mean:


cognitive structure:

the facts, concepts, propositions, theories, and raw perceptual data that the learner has available to her at any point in time, and the manner in which it is arranged.


This definition derives from Ausubel and Robinson (1969, p.51) and White (1985, p.51). Clearly, this definition is somewhat vague without further specification of terms such as ‘concept’ and ‘theory’, which are themselves subject to various interpretations (e.g., for ‘concept’ see Gilbert and Watts, 1983), but, nevertheless, it suffices as a starting point for discussion.

Important features of this construct, cognitive structure, are that:

• it emphasises the importance of the structure (arrangement) as well as the content of a learner’s ‘mind’, to use an even more controversial term! (Caravita & Halldén, 1994; Phillips, 1987);

• it emphasises the dynamic nature of the construct - cognitive structure changes over months and years, and even over seconds;

• it gives a feel for the non-homogeneity of the phenomena: it includes contributions from both perception and storage (memory).

The difficulty of talking about cognitive structure is clear, in that it includes immediate sensory data (with minimal, but of course not zero, ‘processing’), and a range of stored material which will vary along dimensions related to:

• belief - including fundamental commitments, and romanced notions;

• logical nature - from extensive, coherent and well integrated with other ideas, to isolated scraps of ‘knowledge’;

• explicitness - from those ideas that are fully articulated in mind, to the tacit knowledge that effects thinking in ways the knower does not recognise (e.g., Piaget, 1973 {1929} ).

There is little surprise that the literature abounds with so many terms to describe what is elicited in research into the learning of science (Gilbert & Watts, 1983, Abimbola, 1988): alternative conceptions, conceptual frameworks, learners’ ideas, children’s science, gestalts, intuitive physics, p-prims etc. That some of these terms are aimed at the contents of cognitive structure, some at the products of student thinking (a kind of output of processing through cognitive structure), and some at researchers’ models aimed at generalising what has been elicited, adds to the confusion (Taber, 1997a).

In my work I have taken a rather pragmatic approach to this quagmire, which I outlined as:

(1) that concepts are in some way ‘stored’ or represented in a learner’s brain,

(2) and that there is some form of organisation of these representations (i.e. we accept the existence of cognitive structure);

(3) that therefore the notion of two concepts being more or less closely linked, connected or integrated in cognitive structure is a meaningful and sensible one;

(4) that we do not have direct access to a learner’s cognitive structure;

(5) that a learner’s behaviour (statements, responses to questions etc.) may be considered to reflect aspects of his or her cognitive structure;

(6) that we may construct models to represent cognitive structure in terms such as the various conceptions that a learner holds, and how they appear to be inter-related;

(7) that the utility of such models may be judged in terms of the extent to which they are consistent with, and may be used to organise and explain, the learner’s behaviour (statements, responses to questions etc.)

Whilst I imagine virtually everyone in the field would have quibbles with this set of statements, I would hope that, nevertheless, they come close to being acceptable in outline to most researchers. I would consider such a set of statements, or something very similar, as part of the hard core of the research programme in which I am working. The (Lakatosian, e.g. Lakatos, 1970, c.f. Gilbert & Swift, 1985) negative heuristic would suggest that if you strongly disagree with these statements you are not working in the same research programme.

Having said this, there is still plenty of scope for the positive heuristic to operate! Questions such as:

how are concepts stored?

what is the nature of the arrangement?

how is closeness measured?

how is behaviour interpreted to relate to cognitive structure?

etc. provide a rather wide protective belt!

In particular there is a big question regarding the extent to which behaviour, such as answers to questions, reflect underlying aspects of cognitive structure, rather than processing - thinking (c.f. Vosniadou, 1994). Clearly the relationship between cognitive structure and thinking is not simple: the obvious computing metaphor (cognitive structure as hardware, thinking as the running of a programme) breaks down: firstly cognitive structure is dynamic, and, secondly, whereas computer hardware is meant to give - in principle - predictable outcomes, mind has evolved to be highly complex, and is probably better seen as a chaotic phenomena like the weather (c.f. Glick, 1987).

In the present discussion I will assume that it is unproblematic that behaviour (e.g. verbal behaviour) reflects thinking, that in turn reflects cognitive structure: but that the scale at which thinking constructs answers from the resources of cognitive structure is a major research question. There will be a wide degree of variation in the answer: the question of interest here is under what conditions answers are largely pulled from storage preconstructed rather than put together in situ (c.f. Driver et al., 1985; Claxton, 1993).

Taking an overview of the literature, it would seem that there is a significant range of views about the nature of student conceptions that are elicited in research. Some workers focus on the way that elicited ideas may be incoherent, labile, context-dependent, and fragmentary (e.g. BouJaoude, 1991;

Claxton, 1993; Hennessy,1993; Linder, 1993; Russell, 1993; Solomon, 1992, 1993; Viennot, 1979, 1985); whilst other researchers suggest their research indicates student conceptions may be wide-ranging, resilient, theory-like, well-integrated, self-consistent etc. (e.g. Andersson, 1986; Driver et al., 1985; Tytler, 1998; Vosniadou, 1992; Watts, 1982, 1983).

Which view is right? In fact there is no great mystery here. As Smith et al., have suggested,

“A shift toward viewing knowledge as involving numerous elements of different types seems crucial…”

Smith et al., 1993, p.146.

It would seem that anyone undertaking research into students’ ideas (or even spending a little time introspecting their own ideas about fields in which they are not experts: e.g. politics, economics, third world development, modern policing methods…) would surely come to the view that our (human) ideas are indeed often incoherent, labile, context-dependent, and fragmentary. But, similarly, such enquiry will reveal that aspects of our thinking can be resilient, theory-like, well-integrated, self-consistent etc. These descriptors do not relate to absolutes, but to dimensions along which our conceptions may be measured. The vast literature on student conceptions needs to be viewed as indicating under what conditions we are likely to find conceptions with these different characteristics. Unfortunately, some of the literature is flawed by researchers reporting their findings as though an either/or interpretation of student conceptions is appropriate. So whereas many of the careful qualitative studies (such as so much of Driver’s work, e.g. Driver, 1983; Driver & Oldham, 1986) suggest that children’s ideas can be theory-like etc., we find Kuiper suggesting that Watt’s detailed interpretations of student alternative frameworks (Watts, 1983) are invalid because Kuiper’s own normative approach does not not ‘replicate’ Watt’s findings (Kuiper, 1994). There is an obvious point here: answers reflect student thinking, but they also reflect the questions asked (c.f. Kvale, 1996).

From my own reading of the literature, obviously influenced by my own research, I would like to suggest cognitive structure includes the following dimensions:

• explicit knowledge and tacit knowledge;

• general principles of widespread application, and specific data;

• fragmentary knowledge, such as odd facts only weakly related to other data, and extensive frameworks of ideas that may be logically well integrated;

• ideas that are believed with deep conviction, and others to which there is little such commitment: in Kelly’s (1963{1955} ) terms, our constructs may have degrees of permeability;

• aspects for which there is genetic propensity (‘preprogramming’) (Gelman & Markman, 1986; Mithen, 1998; Preece, 1984; Wolpert, 1992), as well as aspects that are derived from physical experience of the world, and from the cultural milieu (Solomon, 1993, 1994).

Figure 1: A model of the learner in the environment

(from Taber, 1997a)

The view of the learner that derives from this (e.g., see figure 1), is of a complex system, taking in information from the environment (including the questions of teachers or researchers), actively thinking about what has been perceived - both at conscious and a sub-conscious levels - and calling upon what is held in memory. The memorised material, has various degrees of grain size including complexes of related ideas (‘conceptual frameworks’), and certain mental tools that act as ‘prepackaged’ units ready to be used as components of new constructions. The genetically determined aspects of brain structure provide filters through which this all occurs: so for example the ‘sensory interface’ puts limitations on what can be seen and heard due to physiology, and our on-going constructions of cognitive structure are influenced by innate factors (represented in figure 1 by the label ‘gestalts’).

That it is possible to ‘store’ ideas in which one does not believe, means that it is not logically impossible to hold in cognitive structure several inconsistent ideas. This enables the learner to hold multiple frameworks of understanding - something that is reflected (although not always expressed in such terms) in many research studies (Ault, Novak and Gowin, 1984; Bloom, 1992; Caravita & Halldén, 1994; Gilbert & Watts, 1983; Hennessy, 1993; Kelly, 1963 {1955} ; Maloney & Siegler, 1993; Mithen, 1998; Pope & Denicolo, 1986; Thagard, 1992, Tyson et al., 1997; c.f. Smith et al., 1993) - whether they are considered as various possibilities, or as my views and my mental model of other views, or as separate conceptualisations that apply in different contexts (Solomon, 1992, 1993).

Bachelard viewed our scientific concepts as having a complex nature (see figure 2), as (in practice) having components deriving from different historical stages (Bachelard, 1968 {1940} ; Souque, 1988; Tiles, 1984), so a scientist’s conception of mass does not fully match the modern physics text-book version (see Taber, 1997a).

Figure 2: Bachelard’s personal epistemological profile for ‘mass’

(redrawn from Bachelard, 1968 {1940} , p.36.)

(from Taber, 1997a)

Results from some studies into student conceptions strongly support a view of cognitive structure that allows such multiple representations (Johnson, 1998; Petri & Niedderer, 1998; Taber, 1995a, 1997a; Taber & Watts, 1997; Tytler, 1998).



Figure 3: A model of alternative frameworks

(from Taber, 1997a)

This deeply synthetic view may seem to be a case of sitting on a fence, and certainly leads to more questions than answers: but any other view would seem to require ‘explaining away’ much of the data in the literature. The brain is a highly complex organ. Mind is a highly complex phenomena. To expect to be able to provide simple models that explain our thinking seems to me over-optimistic, and ultimately counter-productive (c.f. Smith et al., 1993).

Conceptual change.

Cognitive structure is dynamic, it changes over time. Put simply: we learn. (For example, learning has been defined by Petri & Niedderer (1998, p.1075) as “a change in a cognitive system’s stable elements”.) Of course some of us seem to learn more efficiently than others, we all find it easier to learn certain types of things than others, and better in some circumstances than in others, and our students do not always learn the things we were hoping they would!

Given the incredible diversity of ideas about the nature of conceptions discussed above, it is of note that there seems to be a fairly widespread view that learning involves two major categories of change (e.g. Duit et al., 1998; Duschl et al., 1992; Novak, 1985; Posner et al., 1982; Vosniadou, 1992, 1994). In simple terms, there is adding knowledge, and changing what is already known. Of course, unless one holds to a view that knowledge can be totally compartmentalised, any addition will to some extent change the meaning of other stored knowledge (e.g. Ring & Novak, 1971). However, to a first approximation, some additions do not change other meanings in more than a trivial way, whilst other changes actually require the substantial revision of existing knowledge (c.f. Hashweh, 1986). For example, learning that there is an element called ‘boron’ need not significantly alter a learner’s ‘element’ concept, but learning that boron cannot be readily fitted into a dichotomous classification ‘metal-non-metal’ may significantly alter the ‘element’ concept if ‘elements may be divided into metals and non-metals’ is understood to be a key attribute of the concept.

Often Piagetian terminology is used to discuss such issues (e.g. Dykstra et al., 1992; Rowell & Dawson, 1985; Pintrich et al., 1993; Posner, et al., 1982), and we may consider our example thus:

1) ‘boron is an element’ may be assimilated, thus adding an additional example of the element concept (which already includes the attribute ‘elements may be divided into metals and non-metals’), so boron is subsumed under element;

2) ‘boron is a semi-metal, something in between a metal and a non-metal’ may be assimilated as an attribute of the the new concept of ‘boron’ (at this point we might say there is some ‘cognitive dissonance’ - there is a logical inconsistency introduced into the scheme, a ‘dis-equilibration’.)

Accommodation may occur by eliminating the attribute ‘elements may be divided into metals and non-metals’, and perhaps substituting a new attribute of elements: e.g. ‘elements exhibit characteristics along a dimension from metal to non-metal’.

Of course - there are other possibilities. The student could fail to make sense of something ‘in between’ a metal and a non-metal. A second, weaker, meaning for ‘element’ could be opened up (so Boron is an element we learnt about in College, but not a proper element like those - carbon, oxygen, hydrogen etc. - we learnt about in school). Alternatively, the source of the new information may be judged unreliable.

The two-way division between addition to existing cognitive structure, and adjustment of existing features of cognitive structure is clearly a simplistic model, and some researchers emphasise a difference within the latter category between fairly ‘local’ amendments and more drastic restructuring (e.g. Dykstra et al., 1992; Tyson et al., 1997, c.f. Chi et al., 1994).

The importance of context was mentioned above. It is important to recognise that it is the context, as perceived by the student, that is significant. In other words, as Edwards and Mercer have pointed out, context is mental (1987). A learner’s cognitive structure may be viewed as much a part of the learning environment as other factors. Hewson (1985) refers to Toulmin’s notion of ‘conceptual ecology’: that is that the ‘intellectual environment’ provides an ‘ecological niche’ which will differentially support possible conceptual changes. This idea is used by Posner and colleagues (Posner et al., 1982; Strike and Posner, 1985) who suggest that a wide range of features make up such an ecology (anomalies; analogies and metaphors; exemplars and images; past experience; explanatory ideals; general views about the character of knowledge; metaphysical beliefs about science; metaphysical concepts of science; knowledge in other fields; and competing conceptions).

The conditions for conceptual change have been much discussed, and emphasis has been placed on how conceptual change may be considered a rational process (Hewson & Hewson, 1984; Posner et al., 1982; Smith et al., 1993; Strike & Posner, 1985). New ideas must be considered as understandable, plausible, fruitful etc. before they are likely to replace existing ideas. This means that the learner has to construct new representational structures before being in a position to dismiss existing schemes (Nersessian, 1992). The importance of motivational factors has also been pointed out - that other, non-logical, conditions have to be in place for learning to occur (e.g. Pope, 1982, Pintrich, et al., 1993) - although from a Piagetian viewpoint the disequilibrated, perceived-as-inadequate, incoherent state of existing cognitive structure may be seen as motivation enough (Kitchener, 1992, c.f. Dykstra et al., 1992).

The significant question of how any complex, new, idea is ever going to be understood well enough to seem preferable to a cozy, familiar, and often successful existing explanatory scheme has been discussed in depth by Thagard (1992), who considers that we construct alternative frameworks ‘in the background’ until they are well established (and familiar) enough to enter open competition with longer-established ideas (c.f. Schwedes & Schmidt, 1992, p.189; Villani, 1992). It is not necessary for the learner to be consciously aware of such processes, as long as they occur (Caravita & Halldén, 1994; Kitchener, 1992).

One important point is that significant conceptual change is a long-term process, of the order of months and years, rather than the length of the school period or college lecture (Driver & Erickson, 1983; Smith et al., 1993; Pintrich et al., 1993; Stavy, 1998; Vosniadou, 1992, 1994). The literature includes references to aspects of the ‘conceptual trajectories’, ‘learning pathways’ and intermediate states through which conceptual development may occur (Driver, 1989; Duit et al., 1998; Petri & Niedderer, 1998; Smith et al., 1993; Stavridou & Solomonidou, 1998; Vosniadou, 1994). Driver talks of ‘intermediate notions’ “which, though they may not be correct from a scientific point of view, may however reflect progress in children’s understanding” (1989, p.483). Duit et al., refer to the ‘conceptual deviations’ when the ‘sequence of conceptions’ does not lead to the ‘target science conception’ (1998).

The focus of studies into student conceptions in science has, then, shifted from eliciting alternative conceptions, to exploring the dynamics of conceptual development, and the present work should be seen as part of this programme.


The methodology used in this study is described in more detail elsewhere (Taber, 1997a), and will only be outlined here. A grounded theory approach was employed in the on-going process of research-design (e.g. theoretical sampling) and analysis (Charmaz, 1995; Glaser & Strauss, 1967; Glaser, 1978; Taber, 1997b; Strauss & Corbin, 1998). White has commented on how the complexity of the learning process has led to greater weight being put on qualitative approaches such as case studies (1998).

The main data collection technique was the use of in-depth semi-structured interviews. Students studying A level chemistry volunteered to be interviewees in the study, and were characterised as colearners as they were motivated to take part by a belief that they might come to a better appreciation of their own understanding of chemistry (Taber, 1994a). There were 15 colearners who contributed to the study, and most were interviewed several times over a period of an academic year or more. Additional data was collected in the form of Kelly’s repertory test (1994b), recorded dialogues between students, and relevant course work - including concept maps and answers to tests. Case studies were worked up for some of the colearners, including one who contributed a considerable amount of data (being interviewed over twenty times!). A general model was developed by comparisons between the cases.

A range of ‘incidental data’ (Taber, 1997a) was collected from other students - responses to induction exercises (e.g. Taber, 1996a), test scripts etc. - and this material was used as additional ‘slices of data’ with which to compare the findings from the colearners. Key findings from the developing model were used to write two ‘diagnostic tests’, which were administered to a larger number of students to provide a form of validation of the findings of the qualitative analysis (Taber, 1997c, accepted for publication).


The overall results of the ‘Understanding Chemical Bonding’ project are described in more detail elsewhere (Taber, 1997a), and some of the specific findings have been published in the literature (Taber, 1994c, 1995a, 1996a, 1997c, 1998a, 1998b, accepted for publication; Taber & Watts, 1996, 1997). The time available here only allows the presentation of an outline - a brief explanatory model - of the major findings. Each colearner was an individual case with certain idiosyncratic features. Nevertheless, this general model has considerable explanatory power in relation to the data collected. The explanatory model provides a general description which is applicable - in outline - to the different informants in the present study. It may be used as a framework for conceptualising the cases of my own colearners, and by extension as a potentially fruitful starting point for other researchers and teachers wishing to understand and assist conceptual development among chemistry students. (The model arises from this one researcher’s interpretations on my colearners’ thinking. As Kvale (1996) has suggested it is up to the users of interview research to satisfy themselves of the degree to which results are generalisable. However, it is up to the author of a study to provide sufficient evidence for readers to base their judgments. Readers who are teaching chemistry students at a similar level, or undertaking research in associated areas, are referred to the more detailed account in Taber, 1997a.)

The results are presented in the following stages:

A) Some characteristics of student thinking at the start of an A level course are presented;

B) Some areas where students often have difficulty acquiring the curriculum models and explanations of A level chemistry are discussed;

C) Key changes required in student conceptualisation in moving from initial understanding to mastering these curricular topics are suggested;

D) Finally the extent to which students in the study are successful in making these conceptual leaps are considered.


A: characterisation of initial knowledge.

It should be reiterated that each learner’s cognitive structure is considered to be unique (or rather, each learner’s cognitive structure evolves through a progression of unique states), and it is not possible to provide a general description that will match any student in all respects. Nevertheless the research suggested that the following characteristics matched well with the informants in the present study.

1. students commence A level chemistry with a somewhat vague concept of the chemical bond, and more specific (but largely discrete and rather simplistic) conceptions of two bonding categories - ionic (‘electron transfer’) and covalent (‘sharing’ of electrons). Usually the student is aware of metallic bonding, but often has little idea of what this is beyond ‘bonding in metals’.

2. students generally commence A level chemistry with a basic understanding of the electrostatic principles that charge comes in two types, and that similar charge repels and opposite charge attracts. They readily apply these ideas - but not systematically, see below - in some chemical contexts (although the effect of separation of charge is often either unknown, or not brought to mind when considering chemistry problems).

3. although atomic integrity may be related to electrostatic considerations, bonding in general is not explained in these terms. Although electrostatic language may be used to ‘explain’ ionic bonding, closer examination reveals that the student’s concept of ‘ionic bond’ is significantly different to that of the teacher/curriculum. Commonly students tend to adopt a ‘molecular’ framework for interpreting ionic bonding. Sodium chloride, for example, would be seen as made up of ion-pairs that are bound by the transfer of an electron (the ionic bond), with these quasi-molecules typically held together by ‘just forces’ (Taber, 1994c, 1997a, 1997c).

4. chemical bonding is usually explained in terms of a common alternative conceptual framework (Taber, 1997a, 1998a), the octet framework (see the appendix), based around an explanatory principle - the full shells explanatory principle - of the form:

atoms form bonds in order to achieve stable electronic configurations

(variously referred to as octets, full outer shells or noble gas configurations/structures).

5. the octet framework (see the appendix) is used to ‘explain’ chemical reactions, as well as bonding. It appears to be seen as a fundamental and key aspect of chemical theory.

6. students’ conceptualisation of electrical charges and their interactions is often significantly different to the orthodox scientific view, in that the effect (force) tends to be seen as caused by the charge on a body (rather than as an interaction between charged bodies) …

7. … and may be context dependent, in that the interactions in an atom are seen in a particular way: that is nuclear charge (or its effect) is considered shared, equally, between electrons, and the forces acting are selective (the attraction primarily goes to electron from nucleus, and the repulsion between protons is not recognised.)

The common alternative conception that a charged body (such as a nucleus) gives rise to a fixed amount of attraction which can be shared out among oppositely charged bodies is known as ‘conservation of force’ (Taber, 1997a, 1998b).

B: Areas of student difficulty in A level chemistry.

It is suggested that A level chemistry students often have difficulties in learning aspects of the curriculum. (As all teachers will know, there are many such areas of difficulty - here I consider examples which arose from the present research focusing on the concept of chemical bonding).

1. Learners may fail to categorise new types of bond (van der Waals forces, hydrogen bonds) they meet at A level as chemical bonds;

2. Learners may fail to make sense of electronegativity as a continuous dimension (rather than a dichotomy), or to meaningfully understand polar bonds as intermediate between ionic and covalent bonding;

3. Learners may be unable to explain how the structural integrity of lattices is due to bonding;

4. Learners may not accept the stability of compounds that are ‘electron deficient’ or have ‘expanded octets’;

5. Learners may be unable to make sense of patterns in ionisation energies;

6. Learners may be unable to give a deep explanation of why chemical change occurs;

7. Learners may be unable to explain solubility (and deviations from Raoult’s law).

C: Aspects of progression in understanding chemistry.

It is suggested that in order to develop an integrated and comprehensive understanding of chemistry at this level the student must come to

1. see chemical species (i.e. atoms, molecules, ions, etc.) as configurations of charged particles in a balance between attractive and repulsive forces;

2. see bonds as interactions - forces - that are considered to be electrostatic in origin;

3. see reactions as changes in the configurations of charged particles;

4. see chemical stability as a relative notion.

It will be noted that this implies that at A level a key factor is having an electrostatic framework for understanding chemical structures and processes at the submicroscopic level.

D: Students’ developing understanding.

In general A level students’ developing understandings of chemical bonding concern a transition from one conceptual framework - the ‘octet framework’ - that typically characterises student thinking at the start of the course - to another, an electrostatic framework. The latter framework reflects (at least, to a greater extent) orthodox scientific understanding, whereas the former is a common alternative conceptual framework.

In caricature one might suggest that the new A level student explains bonding and bonding related phenomena in terms of the octet framework, and that by the end of the course the successful student has switched to explaining the same class of phenomena using the electrostatic framework. In practice, this would be a vast simplification.


Figure 4: representation of an aspect of cognitive structure of a typical student at enrolment to an A level chemistry course

(figure 12.3 from Taber, 1997a)

So, for example, figure 4 could represent the resources available in memory for a ‘typical’ student commencing A level chemistry. Among various conceptual frameworks present (left-hand box), most of which are not relevant to our present discussion, is a framework of ideas based on some version of the octet rule explanatory principle. This framework, labelled O, is typically the main resource used to interpret and explain phenomena which are perceived as relating to chemical bonding. Most students also have a framework of ideas based around electrostatic principles, here shown as E. This framework is not applied to bonding phenomena as often as O. E is likely to show significant inconsistencies with the version of electrostatics taught in the curriculum: for example some version of the conservation of force explanatory principle (Taber, 1997a, 1998b) is seen to apply in systems such as atoms. Among the concepts that will be held in cognitive structure (and are available to slot into various explanations) are that of the octet rule, covalent and ionic bonding, and atomic shells. The octet rule may well be understood in a fairly orthodox way in itself, although it also forms the basis of the explanatory principle that is the starting point of the alternative framework O. The concept of covalent bond will not have a great deal of sophistication, but will match the ‘expert’ version to some extent, whereas the version of ‘ionic bond’ is likely to be significantly at odds with the curriculum version. The ideas of shells will be the basic concept used to explain atomic structure. If the term orbital is known it is probably seen as synonymous with orbit and shell.


Figure 5: representation of an aspect of the target cognitive structure of an ‘ideal’ student after an A level chemistry course

(figure 12.2 from Taber, 1997a)

By contrast with figure 4, figure 5 represents the corresponding ‘target’ version of this part of cognitive structure. A new framework for explaining bonding has developed, drawing on two different basic sets of ideas: Coulombic electrostatics (C) and quantum theory (Q). C may be seen as a development of E, but is different both qualitatively and in terms of its scope of application. Q introduces the idea of the orbital, and all that follows from this. The joint framework allows the explanation of energy levels, why there are shells (and limits to the number of electrons in a shell: 2, 8, 18 etc.), and a new way of looking at bonds in terms of orbital overlap. In principle much of chemistry can be explained in terms of electrostatic fields, and the restrictions of the quantum rules - for example it is possible to see chemical change in terms of minimising energy due to reconfiguration of charged particles. New concepts may be understood within this scheme (and used as explanatory principles, or intermediate conceptual tools, in constructing explanations and new understanding): new types of bond (hydrogen, van der Waals), electronegativity and bond polarity, concepts of electron spin, atomic spectra, patterns of ionisation energy, molecular shape…

Figure 5 derives from an analysis of the conceptual structure of the topic of chemical bonding, as taught and examined at A level (Taber, 1997a). Figure 6, by comparison, represents the relevant cognitive resources of a particular learner. The subject of this case study was a successful student at A level, achieving the highest grade in the final examination. At the start of his course he was not atypical of the students in the study (i.e., see figure 4).


Figure 6: representation of an aspect of cognitive structure for a successful student at the end of his A level chemistry course

(figure 12.4 from Taber, 1997a)

The two major points to make about figure 6 is that:

(a) it does not closely match the ‘target’ (figure 5);

(b) it does not closely match the initial pattern (figure 4) either.

This student’s understanding of chemical bonding developed considerably during his two year A level course, but it did not reach the intended target structure. Although this student was using electrostatic ideas more by the end of his course, he was still often explaining bonding-related phenomena in terms of the alternative octet framework. This particular colearner also developed and applied an explanatory principle based on minimising energy - that isthat chemical bonds form, and chemical reactions occur, to minimise energy. This principle could be seen as part of an electrostatic framework - but here it was seen as an independent principle, and not related to ideas about electrical fields. For this colearner, the notion of ‘minimum energy’ (which clearly derived from the taught curriculum) was seen as a key explanatory principle in its own right. This meant that the use of the principle seemed arbitrary from an ‘expert’ viewpoint: i.e. explanations were circular, and/or anthropomorphic. However, this did not seem to lessen its power for the student.

Similar results were obtained from other students. Most continued to heavily use octet ideas, although increasingly alongside Coulombic explanations. Quantum ideas tended to cause difficulty (Taber, 1997a), and were not generally well integrated with electrostatic notions.


The results of this research project (as briefly outlined above) will be considered in terms of what they suggest for:

1) cognitive structure;

2) conceptual change;

3) teaching chemistry;

4) the research programme.


Cognitive structure.

The model of cognitive structure used as a guide in this work (e.g. figure 1) is not sufficient to adequately explain all the findings in this research. In particular, data demonstrates the complexity of cognitive structure, that can not be readily represented in simple diagrams.

However, the detailed results of the study leave little doubt that students commencing A level chemistry hold alternative conceptions that can be stable and interfere with the learning planned by the teacher (Taber, 1997a). A particular example is that of a student who held an idiosyncratic meaning of charges ‘+’ and ‘-’ (Taber, 1995a). This student interpreted teaching through a completely different meaning for these symbols (despite the problems this must have sometimes caused making sense of material). Even when the orthodox understanding of charge was learnt, it did not immediately replace the alternative conception. Indeed the more familiar interpretation was often that first brought to mind.

More significant is the alternative conceptual framework, the octet framework, which seems widespread among chemistry students. The research shows that

(a) this framework allows the student to make sense of a limited range of bonding phenomena;

(b) the alternative framework interferes with the intended learning during an A level course;

(c) octet ideas continue to be applied through the two year course, even by ‘successful’ students.

Perhaps, if successful students succeed anyway, one should merely note such a framework as an interesting curio. However, the view taken here is that the successful students could perhaps learn more, and more effectively, if they were not handicapped by starting the course with conceptual goggles (Pope & Watts, 1988) that both distort the material they meet on during instruction, and provide them with a false security that they understand why chemical changes occur. As for less successful students - what can be more confusing than starting an advanced course only to find that what you have already learned does not allow you to make sense of the new work?

This research, then, is not consistent with those who have suggested that ‘alternative frameworks’ are an artifact of the research interview - transient ideas created in situ but having no lasting effect. Nor can ideas like the full shells explanatory principle be seen as an isolated conception which is part of a rag-bag of notions which students carry around. In my research this principle was used over a period of months and years, and was seen to be applicable across a wider range of chemical ‘problems’, and was part of a coherent framework of understanding. Although the octet framework, as presented in the literature (Taber 1997, 1998a), is a researcher’s model, a generalization from a wide range of data, and no individual student was claimed to match the entire framework - it was a model reflecting similar, relative stable, frameworks that were found to be very common among A level chemistry students.

And yet, although I argue this framework was stable and coherent and wide-ranging (or more correctly that these similar individual frameworks, in the minds of individual learners were stable, coherent and wide-ranging), my research also shows that the students learnt to apply alternative, more orthodox, explanatory schemes alongside their octet thinking.

Conceptual change.

I have argued elsewhere (Taber, 1995b, 1997a) that the nature of chemistry is such that it may be profitable to consider the (chemistry-relevant) contents of a student’s cognitive structure to be akin to the tools in a toolbox, where the appropriate tool is selected for a particular chemical tasks. In other words, there may be times when the professional chemist chooses to conceptualise a hydrogen-chlorine bond as ‘a pair of electrons shared between two atoms’, although in a different context a more sophisticated model is needed. More sophisticated versions of concepts may be more scientifically accurate, but there are times when more basic ideas will be adequate and appropriate.

In the research it was found that at least some students actually conceptualise their thinking in this way - that having several alternative explanatory principles to chose from to explain the same concept area is not problematic. For example, being able to explain bonds in terms of octets, or minimising energy, or electrostatic forces was something that did not phase the colearner used as an example above (i.e., see figure 6). He became aware that he had these alternative ‘stories’ on which to base his explanations, and he selected an appropriate (in his view) explanation in each case. Perhaps he would not have developed this awareness if he had not participated in the study - but this at least demonstrates potential for this degree of metacognition in students at this level.

The ‘tools’ in this particular conceptual toolkit did not all match those I was thinking of in my 1995b - some had dubious scientific validity - and the selections this colearner made did not always seem appropriate from the ‘expert’ perspective, but the toolbox analogy does allow us to discuss how alternative conceptions may be both deeply-rooted in cognitive structure, and yet sometimes apparently ignored in favour of alternative explanations. Arguments about life-world versus scientific contexts do not apply here: frameworks based on electrostatic forces or on electronic configurations are both abstract. The debate about whether alternative conceptions are [sic, c.f. can be] stable and significant, or [sic, c.f. as well as], ephemeral and inconsequential can not be settled by simply showing certain conceptions are not elicited in certain (apparently relevant) contexts. Stability in cognitive structure (like in chemistry itself) is relative. An alternative framework may be stable and important, without (completely) blocking the development of alternatives.

In my research octet ideas became less common in students’ explanations as they progressed through their course: but they remained a common feature of most of the colearners’ chemical explanations. Students constructed alternative explanatory schemes based around electrostatic ideas, but perhaps not as quickly as they might, had they not already held their alternative conceptual frameworks in cognitive structure.


Figure 7: conceptual change as change in conceptual profile.

Our consideration of conceptual change, then, is not about when (or if) one scheme is replaced by another, but about the dynamics of a slow change in the degree to which different schemes are used. Mortimer (1995) has suggested applying Bachelard’s idea of a conceptual profile as a tool for exploring conceptual change. Although Bachelard’s profile concerned versions of concepts that related to different levels of epistemological sophistication, we may borrow his representational form to suggest how a student’s explanations of bonding phenomena change through an A level course.

Figure 7 is not meant to be quantitative, but is a graphical representation (c.f. Bachelard, Mortimer) of the changes in the balance between the explanatory principles applied by one of my colearners when explaining chemical bonding and related phenomena. If this diagram is seen as mapping a conceptual trajectory toward a target structure, then this particular - successful - student never arrived at the end of the journey, at least not by the end of his course. (Indeed, further data not yet fully analysed suggests that four years later - having studied a science based subject at degree level - his conceptual profile had ‘regressed’!)

Teaching chemistry.

The present research suggests that A level chemistry courses may be expecting a degree of conceptual change which is unrealistic given the state of students’ cognitive structures at the start of the course.

If the situation stays as it is, some students will continue to be successful in the course, despite tending to explain much chemistry on the basis of dubious science; and other learners will continue to struggle with material that is not compatible with their existing cognitive structure.

One solution would be to change what is required at A level, so that either it requires less conceptual understanding, or much existing material is dropped to concentrate on developing more appropriate central explanatory schemes. That is, either don’t teach material that needs an electrostatic/quantum framework to make sense, or focus the entire course around developing such a framework - and sacrifice whatever does not seem relevant. As I do not expect either development to be seriously considered, I will not dwell on the merits of these ideas.

A further option would be to attempt to avoid students developing the alternative explanatory schemes in the first place. The octet framework is largely inconsistent with scientific ideas - so how does it develop? I suggest this may be partly due to our (chemistry teachers, curriculum designers etc.) sloppiness in presenting science/chemistry in schools (Taber, 1997a, 1995c): i.e. a collective lack of sufficient conceptual analysis of our subject and curriculum.


Figure 8: Alternative conceptual trajectories for learning fundamental ideas in chemistry

(from Taber, 1997a)

Consider figure 8 (which is taken from Taber, 1997a, where it is discussed in more detail). This shows aspects of developing understanding of chemistry. It is suggested that learners need to develop a concept of chemical substance, and to be introduced to a basic particle model of matter before they can understand much chemistry. However, there is still a significant gap (‘A’) between this prerequisite knowledge, and the content of an A level course. What is significant here, I suggest, is the trajectory by which the learner passes across this conceptual chasm.

Route ‘B’ represents what typically happens at the moment. Students learn that in chemistry everything is made from atoms (although this is a dubious point, see Taber, 1996b), and molecules, ions etc. are seen as combinations of atoms. The octet rule heuristic is learnt, and - in the absence of any better available explanatory principle - becomes seen as the main driving force of chemistry (in the form of the full shells explanatory principle). As this principle becomes applied to a wider range of phenomena, a version of the octet framework is constructed. The student now meets A level chemistry - and finds that the existing conceptual structure does not readily fit with much that is now presented in class. The student reacts in the standard ways, re-interpreting the teacher’s words, and selectively learning the new material.

Path ‘C’, in figure 8, offers an alternative conceptual pathway that is designed to avoid these difficulties. In this trajectory there is no discussion of atoms until other material has been learnt. (When particle theory is studied, the term atom may commonly be used, although what is normally discussed is better characterised as molecules!) The learner is first taught about the basic principles of Coulombic electrostatics. At the present time this may be assumed prerequisite knowledge, and implicit in explanations given to students, but this research demonstrates that learners may not share orthodox electrostatics (e.g. Taber, 1995a) , and - in any case - often do not see its relevance to chemical explanations (Taber, 1998b).

Once learners are familiar with Coulombic electrostatics they are then to be presented with an ontology of the basic building blocks of chemistry being charged particles: nuclei and electrons. These entities would be understood in terms of their Coulombic interactions. The next step would be to introduce the orbital concept as restricting the possible locations of electrons in any system so that the configurations that nuclei and electrons take up when they interact are subject to additional constraints superimposed on the electrostatic considerations. (Of course, care would be needed in finding a suitable form of words for such abstract ideas.) This allows some appreciation of ‘shells’, without there being anything magical about ‘octets’ (c.f. Benfey, 1982).

Only then would the concept of the atom be formally introduced! The learner would consider how systems of nuclei, and shells of electrons form, making up ‘atomic cores’. Then further systems of cores and electrons would be considered. An atom is a system of a single atomic core plus sufficient electrons for the charge on the electrons to balance that of the core. Ions and (polynuclear) molecules are other possible systems.

The point here is to see atoms as having similar status to ions, molecules etc., as currently atoms are seen by most students as having a special (unjustified) status. In particular, often, all chemical processes are conceptualised from an initial state of atoms. For example, hydrogen reacts with oxygen because hydrogen atoms need one more electron and oxygen atoms need two more electrons to gain full shells. So two hydrogen atoms share electrons with one oxygen atom. Clearly this is non-sense: hydrogen and oxygen both exist as diatomic molecules, so arguments about needing more electrons to gain ‘octets’ are not appropriate. The ‘assumption of initial atomicity’ (Taber, 1997a) is a key part of students’ octet frameworks. Without this ontological significance of atoms as the basic units of chemistry, students would not be able to develop a widely applicable octet framework.

With a conceptual trajectory like C, chemical change can then be explained as changes in the configurations of cores plus electrons, brought about by unbalanced forces. Bonds may be understood as stable configurations of cores plus electrons that require a significant energy input to disturb them from equilibrium. The learner would then have appropriate knowledge to apply to the content of the A level syllabus.

This scheme has not (yet) been tested. It may prove that such a trajectory is not feasible within the time available for most learners, as the concepts involved are too abstract. (But what does this say about our current approach?) Nevertheless, my recommendations for teaching chemistry would be to try and follow such an approach. In particular teachers should:

• Introduce electrostatics early.

• Avoid over-emphasis of the octet rule, octets, full shells etc.

• Present an ontology based on systems of nuclei and electrons.


The research programme.

This study demonstrates the value of in-depth, long-term, investigation of individual learners’ developing ideas. Only by such studies can questions about aspects of learners’ ideas in science (transient or stable; isolated or integrated; etc.) be looked at in a realistic way. That is, to consider the degree to which different aspects of a learner’s cognitive structure may have these qualities. The full shell explanatory principle was a very common alternative conception that was extremely stable in students’ minds: but it was not always applied in those contexts where it was applicable. As conceptual ecologies change, as alternative explanatory schemes become available, the octet framework was used less.

There are a number of areas of interest that arise from this work. One is the interplay between ideas from different areas of the science curriculum. In the present case, the extent to which students call to mind and apply ideas from physics in chemistry contexts. The present research has shown that - at least for some students - there is a degree of compartmentalisation of knowledge that is unhelpful. This is worthy of further study.

In the present research the focus was on A level studies. It would be interesting to know the extent to which these alternative ideas are retained and applied during undergraduate study of chemistry (or beyond).

It would also be interesting to see how students would cope with the alternative curriculum order suggested above. It might be thought that such an approach would be too abstract and difficult for youngsters to cope with. This could be true, but I wonder if it is any more demanding than the present curriculum? For the expert (teacher), observing the taught material from the higher ground of an extensive and familiar background, our present ways of teaching may make a great deal of sense. However, this research suggests that much introductory chemical theory seems arbitrary and disjointed to our students. They do not have an electrostatic framework to make sense of much of chemistry, so they raise one of the few ‘principles’ they recognise in their courses - the octet rule - to the status of a central explanatory principle. Everything is made of atoms; atoms ‘want’ full shells; accordingly reactions occur and bonds form. Unfortunately once such a scheme is constructed, it has a tenacity that interferes with further learning. It is strongly suggested therefore that teaching along the lines recommended in this paper should be trailed and evaluated.

This present study has other lessons for future research. The similarity between the individual colearners’ alternative conceptions (that they all seemed to have a version of the octet framework) was striking. The widespread nature of the exemplar common alternative framework from physics - an ‘impetus’ framework for force and motion (e.g. Gilbert & Zylbersztajn, 1985) - is perhaps understandable in terms of everyday experience of moving objects in a world of friction and gravity. However, the near ubiquitous occurrence of versions of the octet framework is harder to explain. Research into the origin of this framework could be of value. Three possible avenues to explore would include:

(a) the extent to which the framework - although scientifically invalid - is taught (Taber, 1997a, 1995c);

(b) the role of peer interaction in the social construction of the framework (c.f. Solomon, 1992, 1993);

(c) the possibility that the octet rule ‘taps into’ something basic in cognitive structure (c.f. Watts & Taber, 1996), which provides some form of propensity for the construction of an octet framework.

Whilst being specific to this particular alternative framework, such research could well provide insights of more general value.


Abimbola, I. O., (1988) The problem of terminology in the study of student conceptions in science, Science Education, 72 (2), pp.175-184.

Andersson, A. (1986) The experiential gestalt of causation: a common core to pupils’ preconceptions in science, European Journal of Science Education, 8 (2) pp.155-172.

Ault, Charles, R., Novak, Joseph D. & Gowin, D. Bob. (1984) Constructing Vee Maps for Clinical Interviews on Molecule Concepts, Science Education, 68 (4), pp.441-462.

Ausubel, David P., & Floyd G. Robinson (1971) School Learning: An Introduction to Educational Psychology, London: Holt International Edition (first published by Holt, Rinehart and Winston, 1969.)

Bachelard, Gaston (1968) The Philosophy of No: A philosophy of the scientific mind, New York: Orion Press (original French edition published in 1940).

Benfey, Otto Theodor, The concepts of chemistry - mechanical, organicist, magical or what?, Journal of Chemical Education, 59 (5), 1982, pp.395-398.

Bloom, Jeffrey W. (1992) Contexts of meaning and conceptual integration: how children understand and learn, Chapter 6 of Duschl, R. & Hamilton R. (Eds.), Philosophy of Science, Cognitive Psychology, and Educational Theory and Practice, Albany, NY: SUNY Press, pp.177-194.

BouJaoude, Saouma B. (1991) A study of the nature of students’ understandings about the concept of burning, Journal of Research in Science Teaching, 28 (8), pp.689-704.

Caravita, Silvia & Halldén, Ola (1994) Re-framing the problem of conceptual change, Learning and Instruction, 4, pp.89-111.

Charmaz, Kathy (1995), Grounded Theory, Chapter 3 of Smith, Jonathan A., Rom Harrè and Luk Van Langenhove (Eds.),Rethinking Methods in Psychology, London: Sage Publications, pp.27-49.

Chi, M. T. H., Slotta, D. & de Leeuw, N. (1994) From things to processes; a theory of conceptual change from learning science concepts, Learning and Instruction, 4, pp.27-43.

Claxton, Guy (1993) Minitheories: a preliminary model for learning science, Chapter 3 of Black, Paul J., & Arthur M. Lucas (Eds.), Children’s Informal Ideas in Science, London: Routledge, pp.45-61.

Driver, Rosalind (1983) The Pupil as Scientist?, Milton Keynes: Open University Press.

Driver, Rosalind (1989) Students’ conceptions and the learning of science, International Journal of Science Education, 11 (special issue), pp.481-490.

Driver, Rosalind and Gaalen Erickson (1983), Theories-in-action: some theoretical and empirical issues in the study of students’ conceptual frameworks in science, Studies in Science Education, 10, 1983, pp.37-60.

Driver, Rosalind, Guesne, Edith, & Tiberghien, Andrée (Eds.) (1985) Children’s Ideas in Science, Milton Keynes: Open University Press.

Driver, Rosalind & Oldham, Valerie (1986) A constructivist approach to curriculum development in science, Studies in Science Education, 13, pp.105-122.

Duit, Reinders, Roth, Wolff-Michael, Komorek, Michael & Wilbers, Jens (1998) Conceptual change cum discourse analysis to understand cognition in a unit on chaotic systems: towards an integrative perspective on learning in science, International Journal of Science Education, 20 (9), pp.1059-1073.

Duschl, Richard A., Hamilton, Richard, J. & Grandy, Richard E. (1992) Psychology and epistemology: match or mismatch when applied to science education, chapter 1 of Duschl, R. & Hamilton R. (Eds.) (1992) Philosophy of Science, Cognitive Psychology, and Educational Theory and Practice, Albany, NY: State University of New York Press, pp.19-47.

Dykstra, D. I., Boyle, C. F. & Monarch, I. A. (1992) Studying conceptual change in learning physics, Science Education, 76, pp.615-652.

Edwards, Derek & Mercer, Neil (1987) Common Knowledge: The development of understanding in the classroom, London: Routledge.

Gelman, S. A. and Markman, E. M. (1986) Categories and induction in young children, Cognition, 23, pp.183-209.

Gilbert, John K., and Watts, D. M. (1983) Concepts, misconceptions and alternative conceptions: changing perspectives in science education, Studies in Science Education, 10, pp.61-98.

Gilbert, John K., & Swift, David J. (1985) Towards a Lakatosian analysis of the Piagetian and alternative conceptions research programs, Science Education, 69 (5) pp.681-696.

Gilbert, John K., & Zylbersztajn, Arden (1985) A conceptual framework for science education: The case study of force and movement, European Journal of Science Education, 7 (2), pp.107-120.

Glaser, Barney G. & Strauss, Anselm L. (1967) The Discovery of Grounded Theory: strategies for qualitative research, New York: Aldine de Gruyter.

Glaser, Barney G. (1978) Theoretical Sensitivity: advances in the methodology of grounded theory, California: The Sociology Press.

Gleick, James (1987) Chaos: Making a new science, London: Penguin.

Hashweh, M. (1986) Towards an explanation of conceptual change, European Journal of Science Education, 8 (3), pp.229-249.

Hennessy, Sara (1993) Situated cognition and cognitive apprenticeship: implications for classroom learning, Studies in Science Education, 22, pp.1-41.

Hewson, Mariana, G. A’B. (1985) The role of intellectual environment in the origin of conceptions: an exploratory study, Chapter 10 in West, Leo H. T., and Pines, A. Leon (Eds.), Cognitive Structure and Conceptual Change, London: Academic Press Inc., pp.153-161.

Hewson, Peter W. & Hewson, Mariana G. A’Becket (1984) The role of conceptual conflict in conceptual change and the design of science instruction, Instructional Science, 13, pp.1-13.

Johnson, Philip (1998) Progression in children’s understanding of a ‘basic’ particle theory: a longitudinal study, International Journal of Science Education, 20 (4), pp.393-412.

Kelly, George (1963) A Theory of Personality: The Psychology of Personal Constructs, New York: W. W. Norton & Company (taken from The Psychology of Personal Constructs, first published 1955).

Kitchener, Richard (1992) Piaget’s genetic epistemology: epistemological implications for science education, chapter 4 of Duschl, R. & Hamilton R. (Eds.), Philosophy of Science, Cognitive Psychology, and Educational Theory and Practice, Albany, NY: SUNY Press, pp.116-146.

Kuiper, Jaap (1994) Student ideas of science concepts: alternative frameworks?, International Journal of Science Education, 16 (3), pp.279-292.

Kvale, Steinar (1996) InterViews: An introduction to qualitative research interviewing, Thousand Oaks, California: SAGE Publications.

Lakatos, Imre (1970) Falsification and the methodology of scientific research programmes, in Lakatos, Imre & Musgrove, Alan (Eds.), Criticism and the Growth of Knowledge, Proceedings of the International Colloquium in the Philosophy of Science, London, 1965, Volume 4, Cambridge: Cambridge University Press, pp.91-196.

Linder, Cedric J. (1993) A challenge to conceptual change, Science Education, 77 (3), pp.293-300.

Maloney, David P. & Siegler, Robert S. (1993) Conceptual competition in physics learning, International Journal of Science Education, 15 (3), pp.283-295.

Mithen, Steven (1998) The Prehistory of the Mind: a search for the origins of art, religion and science, London: Phoenix (first published in 1996).

Mortimer, Eduardo F. (1995) Conceptual change or conceptual profile change?, Science and Education, 4, pp.267-285.

Nersessian, Nancy (1992) Constructing and instructing: the role of ‘abstraction techniques’ in creating and learning physics, chapter 2 of Duschl, R. & Hamilton R. (Eds.), Philosophy of Science, Cognitive Psychology, and Educational Theory and Practice, Albany, NY: SUNY Press, pp.48-68.

Novak, Joseph D. (1985) Metalearning and metaknowledge strategies to help students learn how to learn, Chapter 12 of West, Leo H. T., & Pines, A. Leon (Eds.), Cognitive Structure and Conceptual Change, London: Academic Press, pp.189-209.

Petri, Juergen & Niedderer, Hans (1998) A learning pathway in high-school level quantum atomic physics, International Journal of Science Education, 20 (9), pp.1075-1088.

Phillips, D. C. (1987) Philosophy, Science and Social Enquiry: contemporary methodological controversies in social science and related applied fields of research, Oxford: Pergamon Press.

Piaget, Jean (1973) The Child’s Conception of The World (translated by Joan & Andrew Tomlinson), St. Albans, U.K.: Granada (first published in Great Britain by Routledge & Kegan Paul, 1929).

Pintrich, Paul R., Marx, Ronald W. & Boyle, Robert A. (1993) Beyond cold conceptual change: the role of motivational beliefs and classroom contextual factors in the process of conceptual change, Review of Educational Research, 63 (2), pp.167-199.

Pope, M. L. (1982) Personal construction of formal knowledge, Interchange, 13 (4), pp.3-14.

Pope, Maureen & Denicolo, Pam (1986) Intuitive theories - a researcher’s dilemma: some practical methodological implications, British Educational Research Journal, 12 (2), pp.153-166.

Pope, Maureen & Watts, Mike (1988) Constructivist goggles: implications for process in teaching and learning physics, European Journal of Physics, 9, pp.101-109.

Posner, George J., Strike, Kenneth A., Hewson, Peter W., & Gertzog, William A. (1982) Accommodation of a scientific conception: towards a theory of conceptual change, Science Education, 66 (2), pp.211-227.)

Preece, Peter F. W. (1984) Intuitive science: learned or triggered? European Journal of Science Education, 6 (1), pp.7-10.

Ring, Donald G., & Novak, Joseph D. (1971) The effects of cognitive structure variables on achievement in college chemistry, Journal of Research in Science Teaching, 8 (4), pp.325-333.

Rowell, Jack A. & Dawson, Chris J. (1985) Equilibration, conflict and instruction: a new class-orientated perspective, European Journal of Science Education, 7 (4), pp.331-344.

Russell, Terry (1993) An alternative conception: representing representations, Chapter 4 0f Black P. J. & Lucas, A. M. (Eds.) (1993), Children’s Informal Ideas in Science, London: Routledge, pp.62-84.

Schwedes, H. & Schmidt, D. (1992) Conceptual change: a case study and theoretical comments, in Duit, R., Goldberg, F. & Neidderer, H., (Eds.) (1992) Research in Physics Learning: theoretical issues and empirical studies, Kiel: I.P.N. (Institut für die Pädagogik der Naturwissenschraften), pp.188-292.

Smith, John P, diSessa, Andrea A. & Roschelle, Jeremy (1993) Misconceptions reconceived: a constructivist analysis of knowledge in transition, The Journal of the Learning Sciences 3 (2), pp.115-163.

Solomon, Joan (1992) Getting to Know about Energy - in School and Society, London: Falmer Press.

Solomon, Joan (1993) The social construction of children’s scientific knowledge, Chapter 5, in Black P. J. & Lucas, A. M. (Eds.), Children’s Informal Ideas in Science, London: Routledge, pp.85-101.

Solomon, Joan (1994) The rise and fall of constructivism, Studies in Science Education, 23, pp.1-19.

Souque, Jean-Pascal, The historical epistemology of Gaston Bachelard and its relevance to science education, Thinking: The Journal of Philosophy for Children, 6 (4), 1988, pp.8-13.

Stavridou, Heleni & Solomonidou, Christina (1998) Conceptual reorganisation and the construction of the chemical reaction concept during secondary education, International Journal of Science Education, 20 (2), pp.205-221.

Stavy, Ruth (1998) Conceptual development in science education (guest editorial), International Journal of Science Education, 20 (10), pp.1151-1154.

Strike, Kenneth A., & Posner, George J. (1985) A conceptual change view of learning and understanding, Chapter 13 of West, Leo H. T., and Pines, A. Leon (Eds.), Cognitive Structure and Conceptual Change, London: Academic Press Inc., pp.211-231.

Strauss, Anselm & Corbin, Juliet (1998) Basics of Qualitative Research: Techniques and procedures for developing grounded theory, Thousand Oaks,California: SAGE Publications.

Taber, Keith S. (1994a) The teacher-researcher’s duty: accounting for ourselves to our co-learners, paper presented at Accounting for Ourselves, World Congress 3: Action Learning, Action Research and Process Management, University of Bath, July, 1994. (An abridged version of this paper is published in the proceedings, edited by Laidlaw, M., Lomax, P. & Whitehead, J., pp.230-233.)

Taber, Keith S. (1994b) Can Kelly’s triads be used to elicit aspects of chemistry students’ conceptual frameworks?, paper presented to the 20th British Educational Research Association Conference, University of Oxford, September 1994, pp.35-36.

Taber, Keith S. (1994c) Misunderstanding the Ionic Bond, Education in Chemistry, 31 (4), July 1994, pp.100-103.

Taber, Keith S. (1995a) Development of Student Understanding: A Case Study of Stability and Lability in Cognitive Structure, Research in Science & Technological Education, 13 (1), 1995d, pp.87-97.

Taber, Keith S. (1995b) An analogy for discussing progression in learning chemistry, School Science Review, 76 (276), March 1995, pp.91-95.

Taber, Keith S. (1995c) Prior learning as an epistemological block?: The octet rule - an example from science education, paper prepared for presentation at the European Conference on Educational Research, University of Bath, September 1995.

Taber, Keith S. (1996a) Chlorine is an oxide, heat causes molecules to melt, and sodium reacts badly in chlorine: a survey of the background knowledge of one A level chemistry class, School Science Review, 78 (282), pp.39-48.

Taber, Keith S. (1996b) Do atoms exist?, Education in Chemistry, 33 (1), January 1996, p.28.

Taber, Keith S. (1997a) Understanding Chemical Bonding - the development of A level students’ understanding of the concept of chemical bonding, Ph.D. thesis, University of Surrey.

Taber, Keith S. (1997b) Post inductive resonance?: the principles of ‘grounded theory’ applied to chemical education research, paper presented at the symposium on Learning and learning difficulties at the 4th European Conference on Research in Chemical Education at the University of York, September 1997.

Taber, Keith S. (1997c) Student understanding of ionic bonding: molecular versus electrostatic thinking?, School Science Review, 78 (285), June 1997, pp.85-95.

Taber, Keith S. (1998a) “An alternative conceptual framework from chemistry education”, International Journal of Science Education, 20 (5), pp.597-608.

Taber, Keith S. (1998b) “The sharing-out of nuclear attraction: or I can’t think about Physics in Chemistry”, International Journal of Science Education, 20 (8), pp.1001-1014.

Taber, Keith S. (accepted for publication) The truth about ionisation energy: an instrument to diagnose common alternative conceptions, School Science Review.

Taber, Keith S. & Watts, M. (1996) The secret life of the chemical bond: students’ anthropomorphic and animistic references to bonding, International Journal of Science Education, 18 (5), pp.557-568.

Taber, Keith S., and Watts, Mike (1997) Constructivism and concept learning in chemistry - perspectives from a case study, Research in Education, 58, pp.10-20.

Thagard, Paul. (1992) Conceptual Revolutions, Oxford: Princeton University Press.

Tiles, Mary (1984) Bachelard: science and objectivity, Cambridge: Cambridge University Press.

Tyson, Louise M., Venville, Grady J., Harrison, Allan G. & Treagust, David F. (1997) A multidimensional framework for interpreting conceptual change events in the classroom, Science Education, 81, pp.387-404.

Tytler, Russell (1998) The nature of students’ informal science conceptions, International Journal of Science Education, 20 (8), pp.901-927.

Viennot, L. (1979) Spontaneous reasoning in elementary dynamics, European Journal of Science Education, 1 (2), pp.205-222.

Viennot, L. (1985) Analyzing students’ reasoning: tendencies in interpretation, American Journal of Physics, 53 (5), May 1985, pp.432-436.

Villani, A. (1992) Conceptual change in science and science education, Science Education, 76 (2), pp.223-237.

Vosniadou, Stella (1992) Knowledge acquisition and conceptual change, Applied Psychology: an International Review, 41 (4), pp.347-357.

Vosniadou, Stella (1994) Capturing and modeling the process of conceptual change, Learning and Instruction, 4, pp.45-69.

Watts, D. (1982) Michael, Gravity - don’t take it for granted!, Physics Education, 17, pp.116-121.

Watts, M. (1983) A study of schoolchildren’s alternative frameworks of the concept of force, European Journal of Science Education, 5 (2), 1983, pp.217-230.

Watts, Mike, and Taber, Keith S. (1996) An explanatory gestalt of essence: students’ conceptions of the ‘natural’ in physical phenomena, International Journal of Science Education, 18 (8), pp.939-954.

White, Richard T. (1985) Interview protocols and dimensions of cognitive structure, chapter 4 of West, Leo H. T., & Pines, A. Leon (Eds.), Cognitive Structure and Conceptual Change, London: Academic Press, pp.51-59.

White, Richard T. (1998) Research, theories of learning, principles of teaching and classroom practice: examples and issues, Studies in Science Education, 31, pp.55-70.

Wolpert, L. (1992) The Unnatural Nature of Science, London: Faber & Faber.

Appendix: An alternative conceptual framework: the octet framework.

Octet thinking.

The summary of octet thinking, as set out below, is not from any one student, but is a model which has some features that match utterances from each of the colearners. Indeed, in terms of this model, some of the colearners would best be considered to have been in transition between octet and electrostatic complexes at the start of their A level courses. Nevertheless each of the colearners matched the model in some parts.

The core conceptions are:-

• atoms want full outer shells;

• atoms form bonds to obtain full outer shells;

• atoms may form bonds by sharing or transferring electrons;

• a covalent bond is the sharing of electrons;

• shared electrons ‘count’ (for ‘full shell’ purposes) towards the shells of both atoms sharing them;

• an ionic bond is the transfer of electrons between atoms;

• atoms are stable if, and only if, they have full outer shells.

Subsidiary conceptions that may also be present are:

• electrons belong to atoms;

• in a covalent bond each of the bonding electrons is more strongly attracted to its own atomic nucleus;

• when a covalent bond breaks the electrons return to their own atoms;

• in an ionic lattice there is a distinction between the interactions between the specific ions which were formed by a particular electron transfer event, and the interactions between other counter-ions;

• in sodium chloride ion-pairs are (or are like) molecules;

• the true structure of sodium chloride contains two distinct adjacent cation-anion separations;

• the species solvated in salt solutions are atoms, as transferred electrons return to their own atoms as the lattice is broken up.

Octet thinking in chemistry represents an alternative conceptual framework (in the sense of Gilbert and Watts’ 1983 use of the term: “thematic interpretations of data, stylised, mild caricatures of the responses”).

Features of

the octet alternative conceptual framework:

• considering atoms as the basic ontological entities of chemistry, and electrons as parts of (specific) atoms;

• the full shells explanatory principle (that atoms form bonds in order to achieve stable electronic configurations - variously referred to as octets, full outer shells or noble gas configurations/structures) is used as the reason for chemical reaction and bond formation;

• discussing bonding phenomena in anthropomorphic terms, as if atoms were sentient actors;

• imbuing previous states as being significant, in the sense of electrons having ‘memories’ of their origins, and tendencies to act accordingly;

• construing the ionic bond through a molecular model where the bond is defined in terms of electron transfer, and thus the number of bonds is limited to the electrovalency, giving a molecular entity (whether in name or not);

• bonding is construed in terms of the ionic and covalent models that ‘make sense’ according to octet thinking, so that bonds are construed as ionic or covalent (e.g. when they would more appropriately described as polar); or as like ionic or covalent (e.g. the metallic bond may be seen as like an ionic bond);

• interactions that can not be classed as ionic or covalent are considered not to be proper bonds, but just forces.


This document was added to the Education-line database on 31 May 2000