DP Jilid 7, Bi/. 212007
Assessing Conceptual Understanding of Science Ong Saw Lan [email protected]
Introduction The purpose of science education is for students to be able to function in an increasingly scientific and technological world and for citizens who would apply scientific thinking in all parts of their lives. Considered that our lives becoming increasingly complex, we are often confronted with questions which require us to think about and use scientific information so that we are able to make informed decisions. During the last two decades, reform in science education focus primarily on reshaping the vision for science education. Continued effort to improve students' understanding and continuing use of science in their lives remains a major objective of science educators. Coble & Koballa (1996) pointed out that the majority of students graduated from high school knowing science facts without understanding the evolution of ideas and how science affects and be affected by world cultures and societies The practice of science instruction has emphasized of knowing a lot of science, and teachers merely drilling students with scientific knowledge to enable them to pass the tests compromising understanding. Students need to understand the concepts, principles, and theories central to the different science disciplines so that the knowledge acquired is meaningful and beneficial. This will make students view science not as magical or mysterious, but as a powerful and productive means of answering questions about the world and themselves. According to Coble & Koballa (1996), a person possessing understanding should be able to think scientifically and apply the knowledge and skills of science when confronting with both individual and societal problems. The view that 'assessment drives learning' is perhaps the oldest adage in education. The problems of assessment in the sciences are especially acute, partly because there has been such enormous growth in knowledge in this field and because common instructional practices tend to emphasize memorization of science facts with little emphasis on the unifying ideas or concepts of the sciences. Learning science is not about learning facts by rote, but about understanding. Nonetheless, being able to remember certain facts help with the understanding. Science achievement test today does not vary much than in the past when the sole emphasis was on pencil-and-paper, "objective" tests. Methods of assessment need to be change to help to achieve the goal of teaching science for understanding. The purpose of this article is to explain the different levels of understanding that may be attain in the learning of science. First, different meaning of understanding in science is discussed. To provide a different perspective ofthe idea, understanding is further describe at the different levels. To overcome some of the problems in science assessment, some techniques are suggested to help in assessing understanding in science. 27
DP Jilid 7, Bi!. 212007
What counts as conceptual understanding? The term 'concept' means a mental generalization about a class of instances or specific examples. Many concepts can be constructed directly by generalizing from experience of many instances. Often, tough concepts are presented directly and then exemplified. In some cases, instances of concepts are very difficult to demonstrate or make real. For example, it is difficult to exemplify concepts such as 'atom' or 'molecule'. In order that a concept such as 'molecule' may be more readily established, instances are illustrated by constructing representations or models. A model relates to the original it represents only in some ways. By incorrectly ascribing the attributes, this can lead to misconceptions. According to definition of understanding by Newell and Simon (1972), success in problem solving is identify as a critical component of understanding. In school science one of the difficulties of equating understanding to problem solving is deciding when a problem is really a problem, rather than a recall-of-algorithm exercise for students. Conceptual understanding in science involves more than algorithmic exercise. Procedure can be followed without understanding. A more complete definition of understanding science concepts requires the use of qualitative as well as quantitative means to assess understanding. The concept of acceleration, for example, can be defined in tenns of quantitative problem solving (a=F/m, a=v2-vl/t2-tl' etc.) and in relation to prediction of what will happen to objects if they are dropped from an airplane or shot from a cannon. Good (2000) identified three central ways to define understanding. They are explaining science concepts in terms of other concepts, predicting the outcomes of changes in systems, and solving problems that involve more than simple, algorithmic recall. Good (2000) went on to suggest that student's prediction of phenomena in science education can be a powerful way to determine the extent to which a student understands a science concept. Predicting outcomes when a change is made in a system has been a major strategy used in misconceptions research; and problem-solving using the expert-novice model for research has a large data base as well. Each of these pedagogical tools, define understanding in school science in its own way. HarIen (1998) regarded understanding something as having some information about it. Though, information alone is not sufficient. It occurs when one piece of knowledge is linked to certain others to form a concept embracing them all that is usable in a range of contexts. However, it would be possible to have knowledge about the interdependence that was not derived from linking knowledge but was learned by rote. A person with such knowledge could recite a verbal definition but not able to explain or apply it in particular cases. Rote learned knowledge is not usable in forms or in contexts different from that in which it was obtained. In such a case, failure to use the information would imply that their understanding was poor. Harlen(1998) contrasting knowledge learned with understanding by characterizing as knowledge can be transformed, applied in other contexts and used in various ways, such as in making predictions, attempting explanations or making connections between one thing and another.
DP Jilid 7. Bil. 212007
Levels of Understanding in Science Oversby (2002) explores two ways in which science understanding could be characterized at different levels. The first is Bloom's taxonomy of cognitive understanding and the second is a progression in explanations related to models. A modified version of his hierarchy according to Bloom's taxonomy of cognitive understanding, with brief explanations, and examples of the kinds of questions, is shown in Table 1 Table 1: Bloom S Taxonomy of Cognitive Understanding Level Recognition Recall
How concepts can be used The concept has been met before The concept can be distinguished from similar concepts and recalled The concept can be understood in familiar and unfamiliar contexts The concepts can be used in new contexts to develop general ideas The concepts can be used to tease out meaning of a set of data or related concepts The concept is used to integrate a collection of ideas The concept can be compared with other related concepts for the same phenomena for explanatory power
Examples of questions Have you heard of the term 'frequency'? What do you call sound with frequency more than 20 KHz?
Explain what do you mean by resonance.
By considering the forces on a tennis ball thrown into the air, explain its motion from leaving the hand until it hits the ground Explain how you predict a given object will sink or float in water.
By considering all forces acting on an object submerge in water, explain how to determine the upthrust acting on the object Compare the use of renewable energy to that of non-renewable energy sources.
The second type of hierarchical science understanding according to Overs by is known as the ladder of explanations. This approach considers how models are used for explanatory purposes. The difficulties learners have, and the successes they achieve, in understanding science may be understood through a psychological framework of constructing models. The 'model' here means a representation of a variety of targets that can include an object, an idea, a process, a system, an event or an entity. Representations (models) have a variety of expressions, such as physical objects, drawing, animations, and mathematical functions. The mental models of learners also have correspondences and non-correspondences with expressed models as well as the targets.
DP Jilid 7, Bil. 212007
These models have explanatory purposes which appear to progress in the following order (Gilbert et al. 1998): intentional, descriptive, interpretative, causal, involving a mechanism and predictive. Table 2 shows this' ladder of explanations' with a brief explanation about each of the terms and examples of the types of explanations. Table 2: A ladder of explanation (from lowest to highest) Level Intentional Descriptive Interpretati ve
Characteristics of explanation These explanations define boundaries between phenomena Provide information at an observational level Give a way oflooking at reality without intending a cause and effect explanation Cause and effect explanations generally require some generalization across a wide range of phenomena These explanation are built on a wide range of experimental evidence
Techniques for probing understanding
Example Distinguish between evaporation and boiling. Explain changes in temperature when water boils. Using kinetic theory of matter to explain what happen when evaporation occur. Explain why ether evaporates faster than water.
What will happen to a pool of water when then sun comes out after the rain?
of science concepts
There have been a variety of ways of assessing conceptual understanding. Methods use to explore understanding include asking learners to recall information, label a diagram, explain a phenomenon, explain why a particular instance is an example of a concept, or to distinguish between two similar concepts. Those who can distinguish very fine distinctions are thought to show greater understanding. Even though these methods are imperfect, they are in common use in science teaching. The ational Science Education Standards (1996) provide clear information about what should be done in assessing understanding of science. Science tests that assess understanding should focus on application, such as using information contained in a scenario to solve problems, to critique an experimental design, or to make inferences about cause-and-effect relationships. According to Lewis & Linn (1994), the best way for students to improve their understanding of scientific ideas is to test them against their own experience. According to Oversby (2002), concepts can be understood and used at different levels. In order to assess the different levels of understanding, Oversby (2002) considers a variety of ways of assessing conceptual understanding. Problem solving and concept mapping are two important kinds of tests or assessment tools for judging how well someone understands science concepts. Each tool seem to tap an important feature of understanding. Allthough problem solving is seen by most scientists as the ultimate test, concept mapping may be able to give a greater insight into pupils' conceptual 30
DP Jilid 7, Bi/. 212007
understanding than other conventional techniques, For educators interested in assessing students' understanding of science concepts in terms of problem solving, data from studies show how the subject solves or try to solve the problem point out important differences between experts and novices. It was shown that a major difference is the amount of specialized knowledge available to the expert that is not available to the novice. Also, the way this knowledge is structured in the expert seems to be quite different than the way knowledge is structured in the novice. Typically, when defining conceptual understanding in terms of problem solving, objective questions are used to select single correct answers from four or five possible responses. The alternative incorrect responses are often typical errors made by students. The variety of data that can be provided is very informative. Using this data, higher level skills that can then be assessed include (a) Selecting relevant data to form an explanation; (b) Sequencing information to form a coherent and progressive explanation; (c) Choosing explanations at different levels e.g. descriptive or cause and effect; (d) Word or concept association through linking a variety of data. An example of instrument of this type is the well-known Force Concept Inventory (Hestenes, Wells and Swackhamer, 1992) that has been in use over the last 15 years in physics education. The Force Concept Inventory has been used as measures of Newtonian conceptual understanding from high school to university level of introductory physics. The inventory has the advantage of supplying a systematic and complete profile as the design of the inventory recognize the common misconceptions that exist in students' minds on each of the important Newtonian force concept. The alternatives to the items are common misconceptions. Concept mapping is recognized by most science educators as a valid way to assess conceptual understanding besides a useful instructional tool. Concept maps are thought to mirror the structure of knowledge in the brain (Gunstone and White, 1992). The main assumption behind concept mapping is that expertise or understanding can be assessed by asking a person to construct a map by relating concepts in a hierarchical structure using prepositional statements such as is regulated by and results in as the links or connections. The resulting map reflects the person's mental structure related to the concept(s) in question. A variety of methods have been used to enable individuals or groups to construct concept maps. (I) Cards with labels of concepts to be used are given. They are then arranged hierarchically and linked with propositions that are also given. (2) Concepts are arranged hierarchically and propositions devised to link the concepts. (3) The learners can be asked to generate their own concepts Some people have devised ways of scoring concept maps. Points are given for valid links and more points for appropriate hierarchies of concepts. This approach has a number of drawbacks. It focuses on the outcome rather than the process. The product, the concept map, does not represent the group or individual thinking about sensible formats and appropriate meanings of concepts. The map does not reveal the creation of understanding that takes place when the map is being created. A concept map gives us 31
DP Jilid 7, Bil. 212007
much more insight into understanding when we see it being put together. It has much value for the teacher who observes it being created. A concept may be a better formative assessment tool than a summative assessment tool. The third assessment tool or strategy to be considered, prediction taps a unique part of understanding. A student is presented with a system, or a representation of a system, and then is asked how the system might behave if something is done to it. The system can consists of any objects commonly found in school science laboratories that are used to demonstrate certain phenomena related to the science concept(s) in question. Gunstone and White (1992) have described question generating as a tool for assessing conceptual understanding in science. Here, a completion of a question is given and the learner is invited to speculate on what the question(s) might have been. An example is: ... because a metal conducts electricity The form and extent of the questions generated give some idea of the extent of knowledge of the learner.
Promoting conceptual understanding in science through evaluation A number of science educational researchers have suggested that many traditional teaching methods do little to improve students' conceptual understanding. For most science teaching, students and teachers are preoccupied with the acquisition of numerous facts and problem-solving algorithms usually by rote learning. In a study, White and Gunstone( 1992) found that students in a physics course are given a great deal of practice in problem solving skills via problem sheets and tutorial exercises. Inevitably, students concentrate on developing these skills rather than developing conceptual understanding since problem solving skills are the likely one to be tested in the final examination. Other research in science classroom (Tobin & Gallagher, 1987; Mitman, Mergendoller, Packer, & Marchman, 1984) also indicated that teachers are concerned with covering the content in the syllabus and ensuring that students perform well on tests and examinations. These findings present a rather depressing picture of school science education. If this is what actually happening in school science classes, then there is a discrepancy between what is happening and what science educators would like to see happening. In such circumstances it is difficult to see how school science can facilitate higher-cognitive-level learning for most students. In addition, studies of classroom testing have shown that most test items in science do not require more than recall offacts. The consequence is that learning by rote is encouraged as this has help to attain high performance on such test items. Assessment drives learning. As long as the predominant mode of assessment stressed recall of facts and problem-solving algorithms, students continued to engage in rote learning practices. Students thus fail to acquire well-organized conceptual frameworks of scientific knowledge. This misalignment of instruction and assessment in science education discouraged implementation of powerful new instructional practices for meaningful learning.
DP. Jilid 7, Bil. 212007
Conclusion The commonly used assessment practice in science has not been successful in capturing the learning and usage of science knowledge. A wide range of new assessment techniques that do not rely merely on single, quantitative measures of subject matter attainment are needed. Science educators need to invest more time and effort in developing and testing new assessment strategies such as the use of problem solving, concept mapping, prediction and question generation for tapping into the different types of understanding and at different levels. These techniques complement and reinforce meaningful learning that leads to conceptual understanding. Besides, assessing students' conceptual knowledge and preconceptions would be a significant aid to effective teaching. Novak, Mintze & Wandersee (2000) recommended that teachers must plan and implement assessment strategies that support meaningful learning and help to achieve the kind of conceptual understanding that empower students to be more effective in whatever future work they pursue.
References Amos, S. & Boohan. R. (Ed.) (2002). Aspects of Teaching Secondary Science: Perspectives on Practice. The Open University. London Cobble, C.R. and Koballla, T.R. (1996) Science Education in Sikula, 1. Handbook of Research on Teacher Education, Association of Teacher Educators. Harlen, W. (1998). Teaching For Understanding in Pre-Secondary Science in Fraser, 8.J & Tobin, K.G. (Ed.) International Handbook of Science Education, Kluwer Academic Publishers, Boston. Hestenes, D., WeJls, M and Swackhamer, G. (1992). Teacher, Vol. 30,141-158 Mazur, E. (1997) Peer Instruction: A User
Force Concept Inventory.
Prentice HaJl, N.J.
Mintzes, 1.1., Wandersee, J.H. & Novak, J.D. (2000) Assessing Science Understanding: A Human Constructivist View. Academic Press, San Diego NeweJl, A. & Simon, H. (1972). Human problem solving. Englewood Cliffs, NJ: Prentice-HaJl Oversby, J. (2002) Assessing conceptual understanding in Amos, S & Boohan, R. (Ed). Aspects of Teaching Secondary Science: Perspectives on Practice, The Open University. London & N.Y. White, R.T. & Gunstone, R. (1992). Probing Understanding.
Falmer Press, London