Primary Science Teaching - Facts Or Procedures ? Do Different Teaching Approaches Influence

Primary Science Teaching - facts or procedures ? Do different teaching approaches influence children's learning ?

Patricia S. Cavalcante, Lynn D. Newton and Douglas P. Newton

University of Newcastle upon Tyne

This research is partially funded by CNPq - The Brazilian National Research Council

Abstract

Primary teachers with different background qualifications have to teach science in England and Wales according to the National Curriculum. Teachers may influence pupils' learning in science by focusing on different aspects of it. Science lessons were taught to primary classes using three different approaches: process oriented lessons, product oriented lessons and combined lessons. The results show that children's understanding in science was different depending on the lessons' orientation. The combined and product oriented approaches appear to have developed a better understanding of scientific concepts in science. Some implications of this for teacher training and curriculum development are discussed.

Introduction

One aim of science teaching is to enhance pupils' understanding of scientific concepts. Another is to develop children's capacity to understand scientific procedures and to investigate scientifically. Understanding is a mental state, the product of mental processes which infer relationships between elements of information (Johnson-Laird, 1983, 1987; Halford, 1993). Teachers cannot give learners this mental state but can attempt to support the mental processing which gives rise to it. They might do this by providing conceptual structures which embody the relationships or, paradoxically, by withholding such structures (Newton, 1996). However, it would be wrong to characterise these as passive and active learning. Both may require the learner to participate actively in constructing meaning.

The National Curriculum for Science in England and Wales (DFE, 1995) presents a balanced view of science education which encourages both the development of scientific concepts, related to Life Processes and Living Things, Materials and their properties and Physical Processes, and also requires that science teaching should emphasise '... exploration and investigation to acquire scientific knowledge, understanding and skills' and 'use both first-hand experience and secondary sources to obtain information' (DFE, 1995).

Teachers working with Key Stage 2 pupils might, for instance, focus on the subject content of science and develop science skills from these areas of experience. This product centred approach could, for example, give rise to oral explanations and demonstrations of scientific knowledge, and, from time to time, practical activities designed to provide direct experience of phenomena with opportunities to explore and investigate these phenomena. In providing a conceptual structure to help the learner build a functional mental representation, the teacher highlights what is relevant and the nature of the relationships between the elements. For example, the teacher might explain the compressibility of air in a bicycle pump by describing it as dispersed particles which may be brought closer or else by comparing it with the behaviour of a spring.

In contrast, teachers might focus on the processes of science and develop scientific conceptual understanding from it. This process-centred approach could, for instance, offer the children experiments and investigations as starting points for acquiring conceptual knowledge with little or no direct teaching of concepts. In this case a conceptual structure is withheld. The onus is on the children to recall or construct a functional mental representation without reference to a teachers’ description of one. Pupils might infer relationships in the topic under study and may be given an opportunity to test and revise their ideas.

Of course, other teachers might focus on a combination of these two approaches and develop scientific skills and conceptual understanding from in this combination. This mixed approach could be a balance or, perhaps, a compromise, between a product-centred and a process-centred approach, in which the teacher provides a partial conceptual structure and leaves the remainder for children to construct by inferring, hypothesising, or testing their ideas. It could encourage lessons where children do investigations with some features already identified by the teacher, and with some conceptual knowledge about the subject that enables them to appreciate the purpose of the activity. In contrast, it could encourage lessons without a clear purpose which mixed different types of activity, but did not develop either conceptual or procedure understanding exclusively.

We surveyed the science lessons of thirty-three teachers with different backgrounds, teaching older children in the primary school (Key Stage 2: 7+ to 10+ years), and found forms of lessons which exemplified these approaches (figure 1). When introducing a new topic, roughly a third tended to begin by providing conceptual structures. This was often done orally with a demonstration of some phenomenon or event and led to relevant reading and written work. Another third tended to begin by withholding conceptual structures. These usually provided some form of practical investigation of the topic under study. Often, this was preceded by reading task and followed by a writing task which offered a further opportunity to acquire, organise or check relevant conceptual understanding. The remaining third tended to use a combination of an initial conceptual structure and a practical investigation. The latter might involve testing a prediction generated by the conceptual structure or confirming some aspect of it with practical activity. This is not to say that these teachers’ aims were necessarily identical, however, an investigation might be seen as an economical way of providing support for both conceptual and procedural understanding or could be a wish to give priority to the development of skills of enquiry.

Figure 1 - Teachers’ approach for supporting the development of Mental Models

Woolnough and Allsop (1985) suggested that provision for the acquisition of enquiry skills and provision for the acquisition of knowledge of theory should be separate in science if they were to be achieved effectively. Similarly, Killermann (1996) concluded that no single method tested in his research could account for both effective conceptual understanding in science, and high attitude and interest towards the subject. He compared the effectiveness of three teaching methods in Biology education: (a) the children carried out experiments by themselves, (b) the teacher demonstrated experiments to the class, (c) the teacher presented of the same content with no experiment. His results indicated that for 10-12 year old children, demonstration lessons produced more effective conceptual learning in Biology. Children’s scores were significantly higher after this type of lesson, and pupils’ relative increase in problem solving ability was even greater in the demonstration lesson group. He concluded ‘it appears that, at least at this level, demonstration experiments can lead to a considerably greater ability to transfer knowledge and solve problems than experiments that are carried out by the students themselves.’ (p.338).

Duggan and Gott (1995) examined different types of practical work in terms of how they meet the aims of the National Curriculum for science. They found that only the enquiry or illustration practical activities supported the acquisition and consolidation of conceptual understanding. Observation and investigation activities provided the opportunity for children to apply conceptual information rather than learn it. Scientific skills lessons were good for the acquisition of procedural understanding, while investigations supported synthesis and evaluation of procedural understanding in science. Investigation was identified as the most complete type of practical work, because it involved procedural and conceptual understanding. They concluded that, ‘ investigations allow pupils to synthesise their skills, processes and understanding into a overall strategy. It is also only by doing whole investigations, where the data occupy a central role, that pupils will begin to appreciate the significance of the concepts associated with evidence as well as to understand its public nature’. (pp. 146).

However, they agreedthat when the objective was to develop children’s scientific concepts only illustration or enquiry practical activities supported this process.

Driver (1994) has argued that an assumption that investigative and practical exploration activities will give rise to scientific principles and generalisations is naive. She writes that ‘children do make generalisations from first-hand experiences, but they may not be the ones the teachers has in mind. Explanations do not spring clearly or uniquely from data’. For her, understanding in science is supported by the provision of theoretical models and scientific conventions.

On the other hand, Squires (1980) believes that scientific content and the processes of scientific enquiry should not be separated, at least for young children. She prefers an emphasis on scientific investigation as a way of encouraging the development of children’s scientific ideas. Similarly, Feasey (1993) has argued that scientific investigations are a way of enhancing both skills of enquiry and conceptual knowledge simultaneously.

In addition, Screen (1986) explains the process-led approach to science teaching is a good alternative to the content-led ‘overburden’ approach in a sense that it is more stable and readily retained by pupils, and Stohr-Hunt’s (1996) research findings suggest that hands on experiences at least once a week strongly influence science achievement of eight grade students (13 year old children), even when the assessment used mostly focused on conceptual outcomes.

Teachers are often urged to change their approach and to plan lessons like those of others but exhortation is generally ineffective (Fullan, 1991). It may be better to let them know when and how their preferred kind of lesson is useful. Here, our main interest was in the effect of the general forms of science lessons we saw on children’s conceptual understanding. Does it matter if different teachers use different approaches in their science teaching ? Is conceptual understanding supported equally by them ? Does one form of lesson offer an economical way of delivering the different requirements of the National Curriculum for Science ? An experiment was design to gauge the effect of such lessons on older primary school children’s conceptual understanding. These questions were considered in the light of the results and related to practice.

Method

Materials

A number of science lessons on topics of Camouflage, Soils and Materials commonly taught to children between 8 and 10 years were constructed. A presentation by a teacher of conceptual information related to each lesson was videotaped. The presentations information began with a short introduction relating the content of the lesson to everyday life followed by an oral explanation of concepts, a demonstration of the phenomenon under discussion and a brief conclusion. Practical investigations of the concepts were constructed for each lesson, supported by worksheets which offered guidance and questions intended to help children construct the concept in their conclusions. The worksheet required the entry of data in tables, an examination of the results and the construction of a conclusion. Both the teaching of content and the practical activities were designed to reflect tendencies noted in real learning situations noted earlier our survey. The presentations and investigations formed the cores of lessons which began either by providing a conceptual structure (P), or by withholding a conceptual structure (W). Each lesson than had three forms, which reflected the approach used: Product-led (PD-P), Process-led (PC-W), and Combined (C).

In the PD-P form, the lesson provided a conceptual structure and comprised the videotaped presentation followed by an illustrated, textual version of the practical investigation of the phenomenon. The PC-W lesson withheld a conceptual structure and began with an illustrated, textual account of the conceptual information followed by actual engagement in the practical investigation. The Combined (C) lesson comprised the videotaped presentation followed by engagement in the practical investigation, which only partially provide a conceptual structure. All forms thus offered essentially the same information but PD-P, PC-W and C forms differed in how the children were introduced to the conceptual knowledge.

The lessons were tested first on classes of 9 and 10 year old children to ensure that vocabulary, pacing and timing were appropriate. As information was presented orally and in print, it is relevant to note that reading comprehension and listening comprehension are practically identical at this age (Durrell, 1969; Seddon and Pedrosa, 1988; Seddon et al. 1990). All forms of the lesson lasted about 40 minutes.

Test of Understanding and Scoring

The assessment of understanding usually centres on the ability to generate explanations, justifications and, where appropriate, predictions (Johnson-Laird, 1987). Accordingly, a test of the degree of understanding of each topic was prepared for use as a pre and post-test. This comprised open-ended questions requiring explanation and the application of knowledge. For example, for the Camouflage lesson, one of the open-ended questions requiring explanation and application of knowledge in new contexts were respectively:

* There are more black moths in a polluted environment. Why do you think that happens ?

* During the winter in very cold and snowy places, hares change colour and become white. Why do they change colour ?

Lawson et al (1993) devised a scoring system for classroom tests of scientific reasoning, which is intended to reflect different levels of understanding in a subject. It is based on the degree of organisation and the quality of the mental representation indicated by the response to each test question. No representation and irrelevant representation are at the lowest level while a full and detailed functional representation, capable of generating a scientifically acceptable explanation, prediction and justification is at the highest level. Briefly, the scoring system amounts to:

0 - blank, irrelevant remarks or use of given terms without explanation.

1 - misconceptions, explanation based upon various concepts not related to the subject.

2 - partially correct conception plus misconception.

3 - descriptive conception.

4 - partial theoretical conception.

5 - complete theoretical conception.

Random samples of twenty pupils’ responses to each question of each topic were scored according to this system independently by two of the researchers. The indices of agreement (Robson, 1993) ranged from 0.753 to 1.000, with above 0.750 being described as excellent by Fleiss (1981). The Pearson correlation between the scores awarded by raters ranged from 0.83 to 1.00. All pupils’ responses on the pre- and post-tests were scored using this system, with each pupils’ score relating to each lesson being converted to a score out of ten.

Sample

Three parallel classes of 10 year olds were tested for ability in Mathematics and English (Patilla, 1994) and those present for all components of this study were found to be well matched (tables I and II).

Table I - NFER - English Scores

Table II - NFER - Mathematics Scores

Procedure

The test of understanding was applied as a pre-test five weeks before the presentation of the lessons to reduce the likelihood that children might focus attention on particular information when lessons were presented. In consultation with the class teachers, lessons about Materials, Camouflage, and Soils were chosen as these were topics that had not previously been taught. The lessons were presented on subsequent days during one week in the pattern shown in table III. Note that the pattern effectively rotates the approaches through the classes. If the children asked questions, only those to do with the clarification of instructions and the meaning of words were answered. Children worked in groups of 4 to 6 in the practical activities. The children were re-tested immediately after the lesson. All materials were provided by the researchers.

Table III - Experimental lessons' application pattern

Results

The mean scores for understanding increased after all lessons. However, analysis of variance, using the General Linear Model which allows unequal group sizes (Minitab, 1994), showed that some lessons had a significantly greater effect than others (Table IV). The difference between the post and pre-test scores was used as a measure of the gain in understanding. First, there was a large gain in understanding for some topics than for others (p=0.026). Generally, the Soils lessons produced higher mean gains than the Camouflage or Materials lessons. For instance, the effect size for the difference between Soils and Materials in the C form lessons was about 1.00 (Cohen, 1969, describes effect sizes greater than 0.8 as ‘large’). Second, the approaches were not equal in their effect on the mean gain in understanding (p=0.018). Generally, the C and PD-P forms produced greater gains than the PC-W form. For instance, the effect size for the difference between Combined and Process-led/Withholding approaches in the Soils lesson was about 0.95. There was also evidence of a significant interaction between topic (Material, Soils and Camouflage) and approach (PD-P, PC-W, and C). While PD-P approach worked best for Materials, PD-P and PP-C were about equal for Camouflage, and C was the best for Soils (p=0.077).

Table IV - Mean gains in understanding scores for the combination of topics and approach

Discussion

The lessons tested here were not equal in their effect on conceptual understanding. Conceptual understanding seemed best served by the lesson which provided a conceptual structure at the outset. To what extent can this be general ? Lessons are infinitely variable and we cannot say there are no superior lessons which begin by withholding a conceptual structure. However, those tested here were not unlike the lessons we saw being used in schools. A practical investigation preceded by a safety net of relevant textual information and followed by questions encouraging conclusions to be drawn is not an uncommon structure. Yet, even with this safety net, gains in conceptual understanding were generally lower than those produced by the other lesson forms. The results indicate that this particular kind of investigation lesson generally did not provide the best support for conceptual understanding. Of course, the children’s skills of scientific enquiry may have improved, something which is unlikely in the kind of PD-P lesson tested here. If conceptual understanding is seen as a by-product of PC-W lesson, what is gained is useful but less so than that produced by the PD-P lessons.