INTRODUCTION TO THE META-LEVEL FRAMEWORK
In Section Four we described the manner in which the key findings of the literature review were used to develop a framework for thinking about the types of meta-level epistemological development that might underlie children’s observable actions when they undertake science investigations. More specific detail of possible stages in the development of “fair testing” knowledge was also tentatively aligned with each meta-level cluster. Having set out the reasoning process followed, we now present the framework, with some general suggestions for the teaching of investigative skills at each level. We then identify some shortcomings of the evolving framework by outlining components of a science education “for all” (Fensham, 1985) that we think are important but that are not covered within the focus taken here on classroom-based investigative skills.

Section Four described how the framework we have developed draws on various research projects that suggest sequences in the development of various meta-level characteristics of children’s thinking. These include their views of the nature of knowledge, their metacognitive awareness, their meta-level views of the purposes for which activities such as “investigations” are carried out, by themselves and by scientists, and their mental models of causality. It seems to us that some interesting alignments of these various characteristics are possible and we present these as 5 “clusters” below. We have resisted the temptation to call them stages because there is, of course, no certainty that they will develop seamlessly in perfect integrated harmony. Indeed, the complexity of children’s development patterns suggests this is highly unlikely. Nevertheless, the clusters as developed may provide a useful basis for taking account of the range of thinking likely to be present within the diversity of a class of children. Dealing with this diversity in a manageable fashion is the daily challenge that confronts every classroom teacher.

Our attempt to develop these meta-level clusters could prove helpful for teachers in that it provides a relatively simple theoretical framework for thinking about why children do the things they do, and why they face the specific types of learning challenges that their teachers can detect. Such insights are an important aspect of teachers’ pedagogical content knowledge, and help inform in-the-moment teaching decisions. We discuss learning challenges related to the specific teaching of investigative skills, along with potentially helpful teaching foci, for each of the 5 clusters.

CLUSTER ONE: EARLIEST FORMAL SCIENCE LEARNING

The juxtaposition of these characteristics suggests that the very notion of “fair testing” as a formal means of systematically approaching scientific investigations could prove challenging for children whose meta-level views predominantly align with this cluster. Can the relevant investigative skills be actively taught while still keeping open rich opportunities for play, for fostering curiosity, and for extending children’s “library” of knowledge of the material world?

A consistent finding of research in this area is that children can recognise fair tests long before they can produce them via their own planning skills. Accordingly, productive approaches to children’s formal introduction to “scientific investigations” could begin by presenting children with concrete forms of direct “fair testing” comparisons, so that they can see, feel, hear, smell the relevant variables and make direct comparisons between matched pairs of tests.

It is important that the contexts chosen for investigation will yield results that make intuitive sense to young children – that is, the things that happen support and confirm their existing theories of causality. (An example of this is the longer run distances of cars from steeper ramps in the Truck Track task. Children expected this result because of their knowledge of vehicles on hills.)

Teachers could encourage emergent recognition of links between adjacent investigative episodes by using simple visual strategies to record data patterns. Such approaches make no demands on children’s memory space because the “evidence” takes shape before their eyes. We describe one very simple way of doing this for the Truck Track task (and other similar tasks) in Section Six.

CLUSTER TWO: DEVELOPING SKILLS OF EXPLANATION IN SCIENCE

This cluster marks a profound shift from the simple certainties of direct links between causes and effects, and unproblematic knowing about the world. Arguably, the very notion of “fair testing” cannot really make sense until children recognise that there are alternatives between which choices need to be made. There are correct explanations to seek and incorrect explanations to rule out. In this context, a more systematic approach to investigations can yield rewards of knowing for its own sake, and the personal satisfaction of a sense of ownership of new knowledge built.

This type of awareness is not of course limited to fair testing. The ruling out of alternative explanations can begin with exploring, and give shape and direction to pattern seeking and modelling, for example.

Now children can more consistently recognise simple fair tests, and strategies to help them think about these in principle may be appropriately introduced. The ability to retrospectively explain “fair test” choices can be encouraged and children may be making significant steps towards producing their own simple fair tests by managing causal variables.

Again, contexts selected should yield effects that are directly observable and that confirm children’s existing causal theories, so that their confidence in their knowledge building skills can be fostered. Results that can be captured as simple visual and/or categoric data patterns will encourage recognition of these patterns. In such situations, children may be learning to link pairs or short sequences of tests to propose explanations. Possible strategies are described in Section Six.

CLUSTER THREE: A DEVELOPING SENSE OF SCIENCE AS KNOWLEDGE TESTING

Again, this cluster marks a profound shift in the types of ideas and thinking activities children can entertain. Explanation per se will no longer suffice, because various explanations may seem plausible, depending on the theoretical ideas seen to be relevant. With this increasing awareness of the inherent complexity of knowledge building comes the idea of testing personal theories against the available evidence.

Children may begin to experience the power of using the evidence generated by their investigations to persuade their peers to accept their explanations and personal theories. At the same time, they are beginning to compare their own theories with those held by scientists, and to recognise that science ideas are tested against the evidence presented by the world around us. Investigations that involve exploring and/or pattern seeking and/or modelling can all extend these types of theory/evidence links.

Once children can use these types of reasoning skills they might be expected to consistently select fair tests, and to appropriately justify their choices. They might also be expected to accurately produce simple fair tests and hold a short sequence of such tests in the memory while working through them systematically.

As their knowledge of science ideas begins to expand, and their experiences of rich inquiry contexts continue apace, so their personal “library” of cause and effect explanations will also grow. This growth could lead to an increasing ability to identify non-causal variables, and to exclude these from their investigations as they proceed.

The specific types of data gathering employed could be another attribute of increasing the contextual challenge presented by investigations. Students could be using specific types of measurement to begin to collect and process continuous data, although strategies that encourage them to display data patterns as these accumulate will help them keep track of the whole investigative space.

CLUSTER FOUR: ENHANCING SKILLS OF OPEN-MINDED REASONING

There is a clear sense of increasing awareness of the complexity of knowledge building in this cluster of attributes. The sense that the warrants for making scientific knowledge claims should be always open to scrutiny and justification is present throughout the cluster. With this comes the challenge of suspending judgment and considering the various alternatives before deciding on “correct” explanations. The sense of the power of justification of knowledge, through processes that include the consideration and elimination of alternative explanations, can bring aesthetic pleasure.

At this stage students could be beginning to produce more complex sequences of school science investigations that require the management of a number of variables in sequenced tests. Teaching strategies that help scaffold students’ increasing memory capacity will enhance this development.

The choice of investigative contexts in which a number of potentially relevant variables can be easily observed and practically managed will also be essential. The Ball Bounce task has already been identified as having this potential, and we explore it further in Section Seven.

The process of eliminating alternative explanations will challenge students to pay increasing attention to data collection processes and patterns, including sampling protocols, endpoint determinations, measures of variance and central tendency, and the identification of sources of experimental error. There may be an associated recognition of situations where the evidence does not support the formation of a valid conclusion at all.

CLUSTER FIVE: A “KNOWLEDGE PROBLEMATIC” ORIENTATION

How may adults systematically explore the warrants for the beliefs they hold, seeing knowledge as inherently problematic? Only once the attributes of this final cluster are in the process of being achieved can we claim that students might be truly beginning to think as scientists think – at least when the latter are undertaking their professional work. The ability to represent evidence independently of theory as far as is possible has been identified as one hallmark of such thinking.

Again, the qualitative differences between this cluster and those that have preceded it are profound. Here, students are coming to an appreciation of science as a vast nexus of interconnected investigation processes and theories about the world. The appropriate and accurate utilisation of scientific theories differentiates this cluster from cluster four. Such reasoning requires students to hold a broad overview of the framework of key science ideas and inquiry methods in the relevant discipline area, and to know how these connect with the experience(s) at the heart of their own investigations. It is likely that many school students will never reach this level of understanding of the nature of science at all.

This cluster is aligned with a sophisticated view of scientific inquiry, even where the contexts of investigations are necessarily constrained by the equipment that is available to the fledgling scientist. Students working within this framework might be expected to systematically compare different possible combinations of variables to detect and describe different types of interactions. They will be paying closer attention to the quantitative analysis of data outcomes, including the identification and management of measurement errors. A global plan could be used to justify and coordinate a series of activities within one overall investigation.

LIMITATIONS TO THE FRAMEWORK
The framework we have developed, and the literature we have cited, present a traditional, rationalist view of science that is open to both philosophical and practical critique. In this final part of Section Five we can only briefly sketch areas that these critiques might encompass. We do so to signal that we regard this as a work in progress.

The literature selected explores the types of activities that are most likely to be a part of school science investigations, and also some aspects that may be neglected. One of these neglected areas is the focus on theory/evidence interactions (Driver, Leach, Millar, and Scott, 1996) and so they are not usually seen as contentious. However, these interactions arguably should be addressed if students are to come to understand the nature of scientific knowledge building (Osborne, Erduran, Simon, and Monk, 2001). In the selection and shaping of pedagogical material that is at an appropriate level of conceptual challenge for learners of science, the complexity of messy, genuinely new, scientific questions is unlikely to be in evidence. Students could be misled about the intellectual challenge involved in actual scientific research, so that the role of creative thinking could be inadvertently downplayed. Our developing framework makes no mention of creative thinking at present. Whether it should is an open question.

The literature we have reported on the development of scientific inquiry skills usually draws on concepts from the physical sciences that can be demonstrated in relatively simple classroom contexts. Trucks roll down ramps, and boats float down model canals pulled by gravity, for example. This emphasis on simple mechanistic concepts and contexts has been critiqued as misrepresenting science by neglecting the types of systems thinking that are more common in some of the biological and earth sciences (Mayer and Kumano, 1999). What might be involved in the development of systems thinking? Are there missing components that the matrix does not address? Some recent environmental education literature suggests that the scope of the thinking we have outlined should be extended to address this emergent area of research (Sheehy, Wylie, McGuinness, and Orchard, 2000).

If students are expected to be able to transfer “scientific reasoning” skills to other contexts in their adult lives — as “scientifically literate” participants in a democracy — other critiques of the overrepresentation of the physical sciences apply. Zohar and her team (1995) used 4 micro-world simulations that were isomorphic in terms of the types of interactions between variables that were modelled. Two tasks were set in physical science contexts (one being the boat/canal task described above) and 2 were set in social contexts. Interestingly, Zohar reports that participants struggled much more to resolve the social micro-worlds than the physical ones. She speculates that personal theories about social interactions are much more deeply held and so “theory saving” strategies of the type she described are more likely to prevent systematic unravelling of the actual interactions between variables. Do we expect students to develop the skills to make such transfers of contexts? If so, at what stage of development would this be considered appropriate, and what are the implications for curriculum integration? These are questions we could not address.

This report can also be critiqued for the lack of discussion of science as a socially constructed activity/body of knowledge. In the absence of such a focus the only mention of values concerns those that are epistemic and intrinsic to science itself. Values external to science that interact with its knowledge building processes, contribute to considerations of risk and uncertainty, and shape ethical thinking, are not discussed. They are important aspects of the interaction of science with society, so their absence is a significant omission. It seems to us that progression in these areas may also necessitate the development of some cross curriculum links – especially with social studies, and environmental education. Such a project is beyond the scope of this paper. We note that there is some literature in this area (see, for example, Zeidler, Walker, Ackett, and Simmons, 2002) and the inclusion of these aspects within the developing framework would balance the exclusively rationalist focus that has resulted from the project thus far.

One recent research project has suggested that prominent, successful contemporary biologists do not typically derive their hypotheses and research questions from the “pure” theory/evidence thinking of the type that is modelled by the framework. Rather their questions are funding driven and as such, are directly influenced by the funding agency’s social agenda (Abrams and Wandersee, 1995). These researchers found that some of their biologist participants were not even able to articulate the manner in which theoretical thinking in their field had influenced the design of the data gathering technologies they used with such enthusiasm. We have not addressed the nature of interactions between contemporary science and technology at all, and so this is another omission that remains to be rectified.

“Success” in the types of investigations we have described could be critiqued as implicitly presenting a teleological message about the inevitable “rightness” of well-conducted scientific investigations. While the framework does address the epistemological understanding that science knowledge and methods change over time and are always open to revision, we note the absence of the messy history of science stories that present a more contextual view of now-settled science questions. Nor have we addressed the development of children’s ability to understand that different bodies of knowledge may address the same questions from different perspectives and for different purposes. This issue is important as we struggle to find ways of educating children in increasingly multicultural “global” contexts.

Finally, we note that no framework or matrix, no matter how well developed, can present a completely reliable guide for the assessment of the progress of individual children because:

Progression can be defined for a curriculum more easily than it can for assessment purposes. A curriculum, unless we are attempting to create an individualised learning package, can prescribe progression by reference to group behaviour, which is likely to be a little more predictable than that of any individual. Assessment, on the other hand, is by definition personal and subject to all the idiosyncrasies of the individual’s interests, experience out of school, home background and so on, which we know to be so influential in pupil attainment (Gott and Mashiter, 1994, p. 186).

This thought-provoking note turns the focus to the uses to which matrices and/or developmental frameworks may be put. We hope our evolving framework might be found useful for the provision of guidance for planning and structuring classroom activities that can support and extend the investigative skills the students already have. We see dangers in trying to use it as an assessment rubric by which children are judged against normative assumptions of the “stage” they should be at.

We began this section with a discussion of research which demonstrated that children of the same age may be at very different developmental levels with respect to their understanding of the knowledge building processes that constitute scientific inquiry. However, appropriate skilful teaching clearly does make the difference for most children. In Section Six we turn our attention to supporting primary teachers to develop the necessary skills to take up this challenge.

prev page / next page