A class of primary student teachers studied chemistry through modelling as part of their subject specialism in science. The course is aimed at providing a sound pedagogical content knowledge base in chemistry for those intending to be science coordinators in primary schools. Students' prior experiences of chemistry ranged from those taking academic A level chemistry at 18+ to those who finished learning chemistry at 14. Their attitudes varied from anxious and sceptical to a small number who positively looked forward to the course.

Issues such as differentiation, the inclusion of appropriate history of chemistry and the nature of science, and a desire to promote personal and interactive learning, were serious aspects of the teaching and learning process which were considered. A major focus was the development of a metacognitive approach to their own learning on the part of the students.

Learning through modelling was chosen to offer a fresh approach. Modelling in chemistry refers to the processes of representing chemicals and chemical changes through such forms as pictures, drawing, equations, 2D and 3D objects, textual accounts and role play. Great emphasis was laid on the explicit appreciation of points of correspondence between the sources of the models and the targets being modelled, and on points of non-correspondence which should be deliberately ignored when using the model.

Data collection methods included audio-taping of taught sessions, responses by students and tutor and interviews with students, written materials,and a poster workshop assignment. Some of these will be described.

The tutor made a focussed attempt to incorporate research views on teaching about models. Evidence from interviews and a written assignment will be provided as the basis for an evaluation of the success of the method. A widespread commitment to using a modelling approach in the students' future learning was also apparent.

The following account is one of an analysis of part of an extensive research activity. It focuses on only some specific aspects of that activity. The names of the students have been changed to protect anonymity although ethnic origin and gender have been preserved.

Its purpose is to describe a novel approach to the teaching of chemistry to pre-service primary school teachers. This approach was developed to be intellectually challenging, yet supportive of those whose chemical knowledge was relatively limited. The methodology employed was also required to be authentic, that is, to reflect the sort of activity which expert chemists might engage in, and authentic in terms of providing insight into the epistemology of chemistry for those intending to teach it to novices in primary schools.

A model is a representation of an idea, an object, an event, or a system (Gilbert, 1993). A model bears some similarity to the target it is intended to model. In this sense, it exhibits an analogical relationship to the target. Models are created for the purpose of explaining some aspects of phenomena. As a representation, a model exhibits a set of correspondences with the target, but even more significantly, it exhibits a variety of characteristics which do not correspond to the target. Appreciation of the value of a model involves knowledge and understanding of both the correspondences and the non-correspondences. It is usual to be quite explicit about the points of correspondences and, indeed, this is why models are constructed. It is much rarer for any of the points of non-correspondence to be elaborated, and these are either ignored, or felt to be self-evident. It is possible that the creators are clear about the limitations of the models they produce but other users of these models may well not be so clear. There is a distinct possibility that these other users may perceive the models in ways which were not originally intended and may be inappropriate, and could give rise to misconceptions.

The target group were student primary teachers, with specialisms in science, at The University of Reading. Twelve were in the third year, and six were in the second year of their course. Conversations with the students confirmed that there was some apprehension about studying chemistry in detail at this stage.

The following details and comments are representative of the group of students, whose previous experience of chemistry ranged from some study to age 14 years old to being relatively successful at A level, taken at age 18 years old.

Students on the BA(Ed) course study general pedagogy and the teaching of ten subjects at primary level, with more emphasis on mathematics, english and science than other subjects. They also have a significant period of time spent in practical experience in a variety of primary schools. In addition, they all have a specialist subject, which will form the basis of acting as a subject coordinator in a primary school, as well as developing their expertise in this subject at undergraduate level. All the students in this survey study biological and physical science in approximately equal proportions, although most express greater expertise and confidence in the biological sciences.

Since it is impossible to study a wide range of science to undergraduate level in what effectively is one third of a degree course, some selection has to be made. Within chemistry, it was decided that students would focus on areas in which concepts would be developed with primary school children. Since the background was so varied, there was a danger that those with very limited backgrounds in chemistry might be outfaced by a traditional undergraduate chemistry module, and the general lack of mathematical expertise indicated that a quantitative approach could well prove too demanding. Yet, in the light of the need to offer this subject at undergraduate level, the intellectual challenge should be in line with those in the science faculties in the university. The option to focus on developing ideas about chemical change within a modelling framework was a very obvious one to take in the circumstances. Two context were chosen, that of dissolving and that of acidity. This report concentrates on aspects of dissolving, covered in the first four weeks of the ten week course.

In the mind of the tutor, the approach to the topic was as clear as the content. A strong constructivist line was taken, based on knowledge of the sorts of conceptions students may have from the extensive literature on misconceptions in dissolving, and ideas about how these alternative conceptions may be challenged. The examples were chosen to challenge specific ideas which may be present in the minds of the students. As far as possible, the existence of these ideas in this particular sample of students was established through careful questioning. A variety of assessment methods were chosen to investigate both the cognitive development of the students, and changes in their feelings about their learning in the classes. The details of these assessment methods will be reported elsewhere. The focuses for individual sessions were selected to develop areas which experience suggested might cause problems if not addressed earlier. One example of this is the decision to concentrate only on characteristics of liquid solvents in the second session, so that the extent of the interactions with various solutes with these solvents could be explored in depth in subsequent sessions.

The tutor also intended to establish an information rich view about the nature of science, incorporating historical evidence and its related context to indicate how chemical ideas developed, and evidence from recent thinking by academic chemists about dissolving. The students would also be asked to compare explanations of dissolving in teaching texts at levels up to degree study in order to distinguish research thinking about dissolving from the thinking teachers expected novices to develop. In this way, some consideration of the role of the teacher could be made.

Dissolving is a major topic in primary school science in the UK, and a variety of explanations and models are available for study in text books used in schools and undergraduate courses (e.g. in Johnston and Scott (1991), Longden et al (1991), Nusirjan and Fensham (1987), Prieto et al (1989)). Explanations for dissolving are both complex and problematic, and cover both qualitative and quantitative approaches. Dissolving has had a rich history of explanations, culminating as far as school chemistry is concerned with the ionic explanations for salts and other ionic compounds, created at the end of the nineteenth century and the early twentieth century, and by consideration of weak intermolecular forces between dissolving particles, also created at the same time. Explanations of dissolving by academic chemists seem to have been foreclosed in broad terms many decades ago, and this makes it an interesting topic for investigation. Both Brock (1992) and Hudson (1992), in their fairly comprehensive studies of the history of chemistry, provide excellent accounts of the developments of these explanations, making readily available these explanations to anyone who needs them. In addition, there is the opportunity for both repetition of classical experimental studies, and for some modest new experimental material, if required. Explanations of dissolving involve synthesis of existing chemical information and theories, as well as the need to choose appropriately from complementary explanations. The wide range of representations, models, in use, also adds an element of personal preference for the student, sadly lacking from much of chemistry teaching in traditional courses.

1. What modelling in chemistry to a group of primary student teachers with science specialism, but with only modest prior experience of chemistry and with some significant lack of self-esteem, would be motivating, challenging, yet realise worthwhile outcomes in learning chemistry?

2. What pedagogy would be effective in providing a challenging yet supportive learning environment for such students?

3. What are the bases for the tutor's pedagogical content knowledge in implementing a successful chemistry course for primary student teachers?

Evidence was collected by a variety of means. The most common method was the use of audio recordings of the sessions and conversations with students. These were transcribed subsequently. Written evidence was collected in the form of fortune lines, which will be described later, annotated diagrams, examination questions, a poster workshop presentation and transcriptions of recorded explanations made by students. Only some of these will be discussed in detail in this paper. The tutor's (JO) comments in reflective sessions between the two authors were also recorded and transcribed.

Each session occupied three hours with a short refreshment break in the middle. The format was a workshop, with theoretical input sessions by the tutor, practical experiences and group work including discussions. students were encouraged to engage in an interactive mode during most aspects of the course and ask questions whenever there was a need.

Session 1: Fortune lines

Session 2: Discussion of the questions on the annotated diagrams

Session 3: Nature of different solutes

Session 4: Nature of explanation

Further details of session one.

Students were presented with a set of practical demonstrations to promote the modelling process:

1. Crystals of iodine were examined, particularly to note their violet colour. The crystals were then warmed to vaporise the iodine, and the colour again noted. Students were asked to draw pictures to represent what was happening at the particle level. Discussion of the process, in terms of chemical and physical change, was encouraged, focussing on what constituted appropriate evidence. Iodine was chosen as a coloured solute, thus making it more difficult for students to think that the solute disappeared on dissolving. Iodine is a non-polar solute, in contra-distinction to the more common selections of common salt and sugar in elementary classes. Iodine also has the property of dissolving without significant change of colour in some solvents and with a change of colour in other solvents. This difference provides the opportunity for a discrepant event which can challenge thinking about dissolving, and extend thinking about this process,

2. Some crystals of iodine were added to liquid heptane, and the formation of a violet solution of iodine in heptane observed. Students were again asked to draw pictures to represent what was happening at the particle level. Discussion of the process, in terms of chemical and physical change, was also encouraged, focussing on what constituted appropriate evidence. The notions of chemical and physical change were selected as a start point since such an approach constitutes the basis for classifying changes in the curriculum in most schools in the UK.

3. Some crystals of iodine were added to liquid ethanol, and the formation of a brown solution of iodine in heptane observed. Students were again asked to draw pictures to represent what was happening at the particle level. Discussion of the process, in terms of chemical and physical change, was again encouraged, focussing on what constituted appropriate evidence.

4. Students were encouraged to share their pictures with the group as a whole, and to discuss the various contributions.

Fortune lines have been described by Gunstone and White (1992) as a method of studying time dependent affective attitudes. It was decided to use this method in this research to check student enjoyment of the session as it proceeded. With time on the horizontal axis and enjoyment as the positive section of the vertical axis, and anxiety as the negative section, students were asked to brainstorm, individually by writing down their examples, details of how they would be aware that they were enjoying the work, or being anxious about it. They were then encouraged to share their brainstorm ideas by passing round the pieces of paper round the group until everyone had seen all the examples. In groups of three, they were then asked to create three common descriptors of enjoyment feelings and three common descriptors of anxious feelings. In this way, it was intended that there would be some degree of a common approach to constructing the fortune lines, by encouraging agreement about the criteria they would use through discussion. The time for this section was limited to about fifteen minutes and it was felt that this process represented a reasonable compromise in constructing common criteria in the time available. Examination of the completed charts demonstrated a large measure of agreement between the groups concerning th criteria used. The application of the criteria is, of course, subjective, but this is not surprising since the feelings being investigated are also somewhat subjective. Nonetheless, it was felt by both students and tutor that this was a worthwhile exercise.

Students were then asked to reflect on their present feelings of confidence in the session, and to record a point on their fortune line, together with a comment to indicate why they felt this way. They were also asked to leave the line on the desk so that it could be seen by the tutor as he made his way round the room, to give an immediate indication of the feelings of the individuals in the group. At other points during the session, further recordings were made, initiated by the tutor, so that a chart showing how changes in enjoyment or anxiety proceeded during the session. The completed charts were collected at the end of the session for further analysis. In the circumstance, it was decided to plot an average chart for the class, by measuring each point on the chart for each student and then calculating an average. This was quite time consuming but instructive in this session, particularly when taken with the comments made on the charts.

In the first session, the charts showed the following general pattern, with only two significant exceptions:

The two exceptions were one student who claimed the highest enjoyment during the session, and one whose line was similar to the average line displaced downwards so that it started negatively and never reached a positive level.

Comments indicated that they found the session challenged their confidence in their personal understanding of the chemistry behind the process. They only felt that they were gaining some understanding of the chemistry towards the end. They particularly felt that the requirement to share their representations with others in the group was rather threatening.

Fortune lines were used at other times in the course, but an analysis of their value will be reported elsewhere. Their only significance in this report is to validate the tutor's intuitive feeling that adopting a strong constructivist line is discomforting to learners through some thoughtful self-reporting of the students.

An interesting example of how students may become anxious arose during the session. Two students sitting next to each other had drawn similar pictures to represent the solution of iodine in heptane. Both had used the same symbols for iodine, purple circles, and the same symbols for heptane, colourless circles. One had drawn random circles, not touching while the other had drawn pairs of circles, one of each of the purple and colourless circles touching each other. The tutor remarked on the two different pictures but the two students insisted that their pictures were identical. The tutor asked how they were identical and the similar symbols were mentioned. The tutor made the comment that one had the circles not touching and that the other had not, and that one could represent independent particles, i.e. physical mixing, while the other could represent joining, i.e. chemical interaction. At this point, the one who had drawn the particles touching each other said that she had not intended this to be a significant point, that it was not part of what she was trying to represent. At no stage did the tutor make any comment which valued either picture as a better representation than the other. The fortune lines of both students at the end indicated a dip at this point and the annotations said that they were embarrassed at having their work discussed in public, despite the positive approach given by the tutor in the whole incident. This incident only serves as an example of the effects of such incidents on the feelings of individual students. It will be discussed again later in terms of whose representation is being observed, that of the producer or that of the viewer.

A tutor's response to a student has a significant effect on the student's attitude to the lesson. A comment made by a student in one interview was that, although tutors will say that no question is a silly question, their reaction to being asked the question at all, or their subsequent response, indicates that there are some silly questions, at least in the eyes of some tutors. An example of such an interchange took place in the first session:

The tutor has made it possible for the student to ask her question in the first place, by inviting questions and giving students time, in a short silence, to ask something. The student positioned herself as apologetic and ignorant. (Why?)

The tutor sought to put student at ease, by indicating the question would not be ridiculous. He answered it fully, thereby implicitly reassuring the student the question was worth asking. He also indicated and checked with GS, that he was not seeking to position himself as knowing everything. By using the student's name, he also created a feeling that she was a known person in the room and it was OK to ask questions.

Of interest here too, is that, later, JO said to GS that he was surprised by the question, as he assumed that these students would understand the word 'representation' by this stage.

What characterises this chemistry teaching is that the tutor is not adopting the role of the all-knowing expert, but is admitting to some uncertainty, within a framework of considerable expertise in the subject. Communicating this blend of openness while remaining an authority is part of what was an aim of the method of presentation. This was also communicated to the students explicitly, as this interchange demonstrated.

The modelling process is one of representation and it is clear that, in this case, that the representations are considerably different from the target. As part of the modelling process, it is important to consider whose representation it is. In the case mentioned above, where two students appeared to the tutor to have drawn the same thing, but they indicated that the drawing was being perceived in a way which was different ot that which was intended, the tension becomes apparent. If the person producing the model does not make clear how it is to be viewed, including both the correspondences and the non-correspondences, then it is quite possible for the viewer to make up his or her mind about the representation. In the case mentioned above, the model perceived by the tutor was quite different from that created by the student. It is rare for models in text books to be so clarified by the author as to make absolutely explicit the points of correspondences and non-correspondences and, in fact, only some of the correspondences are usually mentioned. Occasionally, even the correspondences are left implicit. The opportunity for misconceptions to arise from such an ambiguous situation is quite apparent here. At one stage even further back, academic chemists often fail to make these points clear so that the role of the model in representing the target has to be inferred, with the dangers for mistakes that this poses.

It was decided to make these points clear to the students at every opportunity as examples of good practice in the modelling process.

The students were asked to draw pictures of the iodine vapour. Most of these consisted of circles scattered at random with large spaces separating the gas.

Here is strong evidence for this student, Janine, appreciating the need to be clear about the points of correspondence and the points of non-correspondence.

In the second session, the students repeated a traditional practical activity of holding a charged piece of plastic near to a jet of liquid issuing from a burette. The tutor decided to use a role play as a model to explain the activity of the liquid in being attracted to the charged plastic. Again, he is very careful to point out both the correspondences and the non-correspondences, as this interchange demonstrates.

The interchange here shows some understanding of the basic features of modelling, on the part of the students involved. There was some evidence that some of the students who had taken part in the role play were not so clear about the explanation it provided. In addition, it is sometimes noticed that learners require some time to accommodate new information and that seeking immediate understanding may not be a helpful way of assessing progress in deep and meaningful learning. Nevertheless, failure to make explicit both correspondences and non-correspondences at the initial learning stage is unlikely to ensure meaningful learning later on.

The final assessment, for credit towards their final degree, was the creation of posters and papers for presentation at a poster workshop. The students were asked to choose an area of chemistry not studied during the course, and research progress in modelling from school level to undergraduate level. They were also required to choose both qualitative and quantitative examples of models, and to identify both points of correspondence and points of non-correspondence in he models selected. Underpinning their examination of the models, they had to provide a theoretical explanation of modelling. All of the students gave at least satisfactory papers and presentations in chemical contexts as diverse as equilibrium, kinetics, and atomic structure. Just under half of these assignments were judged to be very good or excellent in terms of justifying their views about the particular models in theoretical terms.

The students provided good to very good answers to an examination question on modelling (apart from one student who failed to answer any questions satisfactorily). They were able to demonstrate the ability to apply their understanding of modelling to a novel situation (dissolving lithium chloride in liquid ammonia) and to identify the points of correspondence and non-correspondence. In this sense, the course fulfilled its cognitive aims successfully.

Students felt confident in asking the tutor quite deep questions, in the knowledge that they would be treated seriously.

In this account, the tutor is valuing the student's question, and providing quite an extensive explanation in which he makes it clear how he came to his understanding. This explicit form of discourse is a hallmark of his teaching, which appears to engender trust and confidence in his thinking.

We include here a selection of students' perceptions of the tutor's aims in teaching, as written at the beginning of session 2. The purpose of this task was to see if there was a match between his approach and that perceived by his students. It is interesting to see how quickly the students have picked up on the essential approach to the course. Each student's comments are grouped together.

The students have detected very early in their interactions with John, what he values in his teaching and what he feels is worth their learning. They understand that John is not content to just give them information, but in some cases, students feel that scientific knowledge is certain and that therefore there are correct and incorrect answers. Other students already understand that John sees scientific knowledge as tentative. Almost all of them recognise that he is trying to challenge them to think for themselves. They perceive that he cares about them as individuals and that he is interested in their views. We have in this set of student comments, a match between the ways that John conducts his class and students' perceptions of his intentions. We also see that the values dimension of John's pedagogical content knowledge is influencing his teaching and that students perceive his values quite clearly.

The course at the heart of this study has been created to provide a novel approach to learning chemistry for pre-service primary student teachers. It combines both subject knowledge and a pedagogical component, integrated through a focus on modelling.

The analysis has provided some insight into how the tutor combines both his academic chemical expertise with his knowledge about teaching to produce a coherent, challenging yet accessible approach to learning chemistry. Embedded in the learning of chemistry has been the development of confidence among the students in both subject knowledge and in examining their own learning process, no mean feat in a topic which many find to be very difficult.

The aims of the course, in providing in-depth knowledge about dissolving and in promoting interest in the topic has been fulfilled, and attributed, in part, to the tutor's role modelling of enthusiasm, the value of expert knowledge, and making explicit his thinking, during the class sessions.

The study also raises interesting questions about the place of pedagogical content knowledge. Firstly, it provides an alternative route for developing appropriate chemical pedagogical content knowledge for primary teachers, other than the more normal route of presenting factual content as unproblematic. Secondly, it exposes some of the pedagogical content knowledge of the tutor, and asks questions about the sort of knowledge which tutors should have in preparing novice teachers.

1. The process of modelling can be motivating and fulfilling for some students learning chemistry, while remaining appropriately intellectually challenging.

2. Modelling requires both points of correspondence and non-correspondence to be identified if the viewer is to appreciate the model in the same way as the provider. If this is not done, then what may be perceived may simply be a new model created by the viewer, and this could lead to misconceptions.

3. Students may not appreciate a sufficient range of models, including those in role play, for example. Care needs to be exercised in making explicit the nature of the models in each explanation.

4. The confidence of learners has a significant impact on their effectiveness of learning. Fortune lines can provide a readily available and open method of recording confidence and similar feelings.

5. Teachers may be more effective in their teaching if they act as role models of learning in their teaching. Teachers may need to give more attention to a metacognitive understanding of their position of authority, while being prepared to share uncertainty from a generally expert position.

6. The effectiveness of incorporation of the history and philosophy of chemistry in promoting deep learning has been indicated and warrants further consideration.

There are many avenues for further research suggested by the research given. These include: experiments to teach modelling to other learners in other contexts, schools, universities and colleges; appreciation of the modelling process by teachers; delineation of points of correspondence and non-correspondence for a range of models to provide case studies for consideration. It may well be possible that discussion by learners of specific models may prove both an efficient method of formative assessment as well as giving access to how the learner understands theories associated with the models. The study raises questions about the pedagogical content knowledge of both primary teachers, and their teacher educators (tutors). What would count as appropriate is a fundamental issue. How teachers and tutors could readily gain access to the history and philosophy of science is also worth exploring further.

Brock, WH (1992) The Fontana History of Chemistry Fontana Press (London)

Gilbert, JK (ed) (1993) Models and Modelling in Science Education Association for Science Education, (Hatfield)

Gunstone, R and White, R (1992) Probing Understanding Falmer Press (London)

Hudson, J (1992) The History of Chemistry Macmillan, Basingstoke

Johnston, K. and Scott, P (1991) Diagnostic teaching in the classroom: teaching and learning strategies to promote concept development in understanding about conservation of mass on dissolving. Research in Science and Technology Education 9 (2) 193-212

Longden, K, Black, P and Solomon, J (1991) Children's interpretation of dissolving. International Journal of Science Education, 13 (1) 59-68

Nusirjan and Fensham, P (1987) Descriptions and frameworks of solutions and reactions in solutions. Research in Science Education 17, 139-148

Prieto, P, Blanco, A and Rodriguez, A (1989) The ideas of 11-14 year-old students about the nature of solutions. International Journal of Science Education, 11 (4), 451 - 463