This paper presents results of research that explored students' constructs about energy During the implementation of a constructivist learning approach. Results from repertory grids showed that students who experienced the construetivist approach provided significantly more constructs, and of a wider nature, than students who were taught in their usual fashion.
This paper presents results of research that explored students' constructs about energy During the implementation of a constructivist learning approach. Results from repertory grids showed that students who experienced the construetivist approach provided significantly more constructs, and of a wider nature, than students who were taught in their usual fashion.
This paper presents results of research that explored students' constructs about energy During the implementation of a constructivist learning approach. Results from repertory grids showed that students who experienced the construetivist approach provided significantly more constructs, and of a wider nature, than students who were taught in their usual fashion.
Research in Science Education, 1999, 29(4), 515-525
Students Constructs About Energy and Constructivist Learning
Tony Fetherston Edith Cowan University Abstract This paper presents results of research that explored students' constructs about energy. During the implementation of a constructivist learning approach, eoustructs from students in three classes were elicited using a repertory grid method, both before and after completing a module of work dealing with energy. Results from repertory grids showed that students who experienced the construetivist approach provided significantly more constructs, and of a wider nature, than students who were taught in their usual fashion. Results from a school test administered to all three classes and the differences found between the results from the repertory grid techniques and the school test are then discussed. This paper concludes that students taught with the constructivist approach had much increased personal knowledge concerning energy, more so than those students taught in the traditional manner, and also that students taught with the eonstruetivist approach learnt the school science equally as well as those students taught in the usual fashion. Over the last t went y year there has been an increasing amount of international research conduct ed by science educators dealing with the learning of science. Constructs like misconceptions, chi l dren' s science, conceptual change and eonstructivism have all been applied in this area. The research has been summadsed by researchers like Osborne and Freyberg (1985), Driver (1989), and Northfield and Symington (1991) as significant historical examples. Essentially this large body of research regards the learner having pre-existing ideas about some of the phenomena that t hey encounter in science lessons and bringing these ideas to bear upon lessons t hey encounter. This body of research has also shown that children' s ideas are usually firmly held, difficult to change but unfortunately often scientifically incorrect. Almost all o f the results of this research can be encompassed under a constructivist view of learning. Most educators operating within the construetivist paradigm would agree that before learning can occur a student has to engage with the processing of the perceived information. Most woul d also agree that knowledge is constructed by a student using cognitive processes that are not necessarily conscious. This active processing of information is assumed to occur with the presence of recalled pri or knowledge. A constructivist view of learning accepts that initial constructed understandings can be very limited in scope and that the building o f conceptual structures requires effort and purpose on behal f of the learner. Taking a longer historical view, constructivism can probably be connected to the writings of Plato and probabl y more recently to Kent, Giambattista Vico, Froebel, and in this century to Dewey (Watts, Jofili, & Bezerra, 1997). More recently, over the last two decades different researchers have identified different forms of constructivism. For example, StelTe and Gale (1995, p. xiii) distinguish six different core paradigms: social constructivism; radical constructivism; social constructionism; information-processing eonstructivism; cybernetic systems; and sociocultural approaches all of which "reject traditional epistemological claims about knowledge as an objective representation of ' real i t y' " (p. xiii). The most commonl y discussed of the above six types are radical and social constructivism. Radical constructivism concerns itself mostly with knowledge of the experiential kind of which yon Glaserfeld is most commonl y associated. This view of constructivism woul d contend that validation of a per son' s knowledge occurs through the extent to which their evol vi ng fits the 516 FETHERSTON person' s experience. Radical constructivism tends to have a focus on individual self regulation, similar to a Piagetian view, and the building of conceptual structures. This view is concerned with mostly cognitive processes. Social constructivism embraces the role of language in meaning making and the legitimisation of knowledge. In this form of constructivism the socially and culturally situated nature of mental activities is of prime importance. Vygotsky could be termed a social constructivist and more recently in science education, the work of Rosalind Driver would be mostly associated with this kind of constructivism. No matter what label we place on the particular flavour of constructivism, most would agree that people actively construct knowledge and that social interaction can assist the process. Perhaps the main distinction between these two large groupings above would be the loci of study: "In radical constructivism, the focus is cognition and the individual; in social constructivisrn, the focus is language and the group" (Staver, 1998, p. 504). Attempts have been made to integrate these two different perspectives (e.g., Cobb, 1994). This socially constructed dichotomy is useful in enabling us to construct our own understandings about how students learn. However not all versions of constructivism fit neatly into this scheme or the six listed earlier. For example, critical constructivism, which is concerned with "constructs which entail knowledge in the world fast-hand experience, 'hands-on," on-the-scene understanding of phenomena" (Lowe, 1982, as cited in Watts, Jofili, & Bezerra, 1997) may be difficult to classify with this scheme. Personal Construct Psychology Similarly Kelly' s (1955) Personal Construct Psychology (PCP) also does not fit neatly into the above categories. One of the common criticisms ofeoustructivism in general is that it does not have a sound epistemologieal base (Mathews, 1994), despite its general acceptance as an excellent pedagogical strategy. This criticism cannot be levelled at PCP as it is an explicitly stated, well articulated theory founded on a fundamental postulate and 11 corollaries which provides a clear theoretical framework upon which researeh can be based and to which findings can be referred. Kelly' s theory emphasises each individual's unique construction of the world but in contrast to Piaget, provides a theory that emphasises the social nature of such constructions. An important emphasis in PCP is on communication and the sharing of meaning and this psychology integrates affective and sociocultural factors with cognitive factors. Kelly' s epistemology and theory of personal constructs allows the distinction between personal meaning and the formal knowledge of science to be bridged. Many of the corollaries are directly applicable to current science education research and can provide explanatory power to these mostly constructivist studies. For example, the fragmentation corollary states that "A person may successively employ a variety of construct subsystems which are inferentially incompatible with each other." When view in the light of this corollary, misconceptions, alternative frameworks and the like are seen as a natural and normal part of every student's construction of science ideas. It is a well-developed eonstructivist theory and provides a sound basis for a constructivist approach to science learning. It has the potential to guide teachers in how best to assist students to construct science knowledge. Practical Approaches It is generally accepted by most researchers that there are sound educational reasons for adopting constructivist learning approaches. These include: the ability to address students' own ideas; the ability to bring about meaningful change in students' often incorrect science ideas; and E N E R G Y A N D CONSTRUCTIVIST LEARNING 517 the increased potential for students to construct rich understandings of science concepts. Various constructivist science teaching models have been developed. Researchers such as Driver and Bell (1985), Osborne and Wittrock (1983) and Cosgrove and Osborne (1985) have all proposed practical constructivist learning models. Conceptual change research, exemplified by Posner, Strike, Hewson and Gertzog (1982), could also be classified as constructivist in origin. Accepting the potential of constmctivist learning for improved educational outcomes and using Driver and Bell (1985) as a guide to development, a learning approach was developed by the author (Fetherston, 1997) grounded in Personal Construct Psychology (Kelly, 1955). The practical approach developed was cyclical in nature with the starting point being the determination of the teacher's and the student's existing conslructs. The approach had a focus on individual constructions and the sharing of meaning with conversation and group work an essential part of the model. The theoretical underpinnings of the model, its development and validation are explained in detail in Fetherston (1997). Design and Methods During the implementation of this approach, a variety of methods were used to probe students' understandings of the topic energy. Among the methods used were Interviews-about- Events (Osborne & Gilbert, 1980), Questions-About-Events (Fetherston, Submitted) and observation of students and teacher during lessons. Apart from these techniques the author decided to also use a repertory grid technique (Kelly, 1955) as a means of determining students' constructions about energy. This decision was based on results from preliminary work (Fetherstonhaugh, 1994) which demonstrated the validity of the approach and its capacity to reveal students' constructs about energy. The technique was also derived from the theoretical underpinnings (Kelly, 1955) of the learning approach developed. Repertory grids can yield from individuals data that have both breadth and depth. Unfortunately in a classroom containing 30 individuals, the technique is too time consuming to be administered through individual interviews, which is the usual method. Consequently a paper and pencil version of the repertory grid was constructed using supplied elements, an accepted repertory grid technique. The following supplied elements were used in this study: solar energy; electricity; energy from food; energy from coal; nuclear energy; energy in a moving bullet; stored energy; energy from chemicals; and heat energy. Students were asked to compare the above elements three at a time (triadic elicitation) using the standard repertory grid question "In what way are two of these elements the same but different to the third?" The answer to this question (the emergent pole of a construct) was written on one end of the supplied grid sheet and the opposite of this pole was written on the other end of the grid sheet. Comparisons between the elements proceeded until no more constructs could be elicited. Readers requiring more explanation of this technique are referred to Fetherstonhaugh (1994) or to Fransella and Bannister (1977) for further detailed explanation of repertory grid technique. Constructs from students involved in the study were elicited immediately preceding the implementation of the learning approach and again immediately following the conclusion of the implementation. Prior to implementation, students were given a short introduction to the technique and elicitation proceeded smoothly with students having few problems. 518 FETHERSTON Research Question The purpose of the study was to determine if: (i) there were differences in students' constructs about energy between students taught with traditional methods and students experiencing a particular constructivist learning approach; and, (ii) if differences did exist, what was the nature of the differences. Sample The study involved three classes of Year 9 students attending a randomly selected inner suburban senior high school in metropolitan Perth, Western Australia. Year 9 classes were chosen at this school as the topic energy was normally taught at this time to these students. Following quasi -experimental techniques, the treatment (the constructivist teaching approach) was randomly assigned to two of the groups and the other group were taught in their normal fashion enabling comparisons to be made between classes. Students in these classes were allocated randomly at the start of the school year into classes similar in range of ability and came mostly from a middle to upper socio-economic area. The school had an academic orientation with high numbers of students eventually completing five years of secondary schooling. The classes in this paper are identified by their teachers' pseudonyms, Scan, Rob and Rick. Scan and Rob implemented the constructivist approach in their classes and Rick' s class studied the topic in its usual traditional manner (see below). All three teachers had at least l0 years experience and were regarded by their peers as very good teachers. Classes contained boys and girls and there were 31 students in Sean' s class, 33 students in Rob' s class and 30 students in Ki ck' s class. The topic of study was an introductory unit dealing with the topic of energy. The traditional approach in this school was characterised by these teachers involved in the study usually using a common programme which ensured a uniformity of coverage of content amongst their classes studying the topic and involving generally uniform assessment. Hands-on practical work and teacher led discussions were the main approaches taken. Lessons usually involved some combination of the following strategies: practical sessions; teacher led discussions; book work; video wat chi ng and discussion; and library research sessions. The general approach was transmissive in nature with the outcome that students finished the topic with a set of notes which comprised summaries of important scientific ideas, calculations, results from practical activities and photocopied worksheets. This set of material comprised the content to be learnt for the topic test and students were helped in this task by the teacher conducting revision of the material close to test time. All three classes covered the same specified content and spent the same time on it. Analysing the Data Data from the repertory grids, being basically a matrix of numbers, can be analysed statistically in many ways, for example by calculating ranges, mean values, standard deviations, correlations, by ANOVA techniques, principal components analysis or duster analysis. However being essentially a qualitative, individualistic method, repertory grids elicited from groups of individuals require techniques which enable the researcher to gain valid measures of group construing. In this study, a method was used which established common groups of constructs. The method was based on the assumption that the person doing the grouping (the author) attached the same meaning to the verbal labels of the constructs as did the respondents. This grouping process was used for students' constructs that were elicited before and after implementation and elicited ENERGY AND CONSTRUCTIVIST LEARNI NG 519 constructs were not difficult t o assign t o the broad groups established. Constructs were assigned into groups based on the first (called emergent) pole of the elicited construct. The grouping process was repeated for each set of grids from each class before and after implementation. Those groups of constructs established whi ch had four or more constructs, representing a construct held by four or more students in a class, are reported. In this way a composite picture of each class' s construing was established based upon the students' own views of energy wi t hout resorting to sophisticated statistical techniques. Results Results are reported for each class, for conslructs elicited before and after the implementation of the learning approach, in the two classes undergoing the constructivist approach (Sean and Rob) and for the class that was taught in its usual way (Rick). Constructs Elicited from Classes Before Implementation of the Learning Approach Using the grouping procedure described above, Table 1 was constructed. Interpreting this table, it can be seen 20 students in Sean' s class wrote a construct on their sheet that could be classified into a group called Natural, based on the verbal label of the first (emergent) pole elicited. Table 1 lists the ideas brought by these students to their science lessons and represent the ideas that most constructivist approaches woul d seek to engage and deal with. Table 1 Constructs Elicited from all Three Classes, Pre-Implementation N u m b e r o f constructs N u m b e r o f constructs N u m b e r o f constructs C o n s t r u c t g r o u p f r o m Scan's class f r o m R o b ' s class f r o m Pick's class (Total N=126) (Total N=126) (Total N= 126) Natural 20 18 15 For heating 13 6 6 From sun 10 Dangerous 12 , 4 5 Chemical 7 7 Stored 6 13 5 Renewable 6 Fast 6 Manufactured, mechanical, 5 15 processed, activated by man Strong/Big/Solid 5 Widely used 5 5 From coal 5 Used in living things 8 11 U~ed by people 4 Environmentally clean 5 Used at home 4 8 Pollute 4 Immediate 4 From Heat, Explosion or Burning 4 Provides energy 4 Other 14 11 13 520 FETHERSTON The most common consmacts in Sean' s class were those that could be classified under a label relating to Natural, Heating or Dangerous. In Rob' s class students viewed energy in terms of whether it was Natural, Stored or Originally from the sun. The three most popular constructs in Kick' s class were grouped under headings called Natural, Manufactured, mechanical, processed, activated by man and Used in living things. For heating was also a construct appearing frequently i n all three classes. The constructs Natural, Stored, For heating and Used i n l i vi ng things are common t o all three classes (with many others common to at least two classes) and students' views of energy, examined on a class basis, centre around these common constructs with a few additional constructs, held by small numbers of students in each class. This means that when students in these classes encounter formal science ideas about energy in lessons they are most likely to see these new ideas in terms of their existing constructs: they will relate new ideas about energy to whether the energy i s Natural, Stored, Dangerous or Used in living things. Such constructs can represent common growth points for new ideas in students in these classes. Students' ideas about energy, pre implementation in these classes, are not diverse and show little evidence of formal science ideas. Nonetheless there are some differences in the most common constructs between classes. This indicates t hat there is variation in the ideas that students bring to their science classes, most probably arising from their different formal science backgrounds, as initial class allocations were random. Using the identical grouping procedure described above, Table 2 was constructed using constructs elicited after the implementation of the learning approach. From Table 2 (post implementation) two features are immediately obvious. In the classes that underwent the constructivist learning approach both the absolute number of constructs and the number of construct groups had increased markedly whereas in Kick' s class, which experienced their usual teaching, there was little difference in either number or groups of constructs. Interestingly the popular (pre implementation) construct Heating did not appear after instruction. This broad construct may have become more refined resulting in it being categorised under the heading Associated with heating. Table 3 shows that there is a much larger mean number ofeonslructs per student elicited from each student in Rob and Sean' s classes, after implementation, compared to Ri ck' s class. The differences in numbers of constmets between classes were statistically significant (F(5,170)=84.4, p<.0001). Scheff~ tests were used to test for differences between groups and showed t hat there were no statistically significant differences between the classes in the number of constructs elicited from each student prior to implementation but there was a significant difference in t he number of constructs elicited from each student, after implementation. Students in Scan and Rob' s classes, who used the constructivist approach, gave statistically significantly more constructs per student than students in Kick' s class. The difference between Rob and Sean' s classes was not statistically significant, after implementation. All classes showed an increase in the number of constructs elicited after implementation. Kick' s class reported an increase of 1.72 in the mean number of constructs elicited, per student, which was not statistically significant. Rob and Sean' s classes reported an increase of nearl y 11.37 and 9.79 constructs per student, respectively, differences which were statistically si gni fi cant (dr< .01). ENERGY AND CONSTRUCTI VI ST LEARNI NG 521 Tabl e 2 Constructs Elicited from Classes After Implementation of the Learning Approach N u m b e r o f constructs f r o m N u m b e r o f constructs N u m b e r o f constructs Construct g r o u p Scan's class f r o m R o b ' s class f r o m Pick's class (Total N--434) (Total N=452) (Total N = 188) Natural 25 (23) * 21 18 For heating From sun 23 32 (26) * 7 Dangerous 16 16 I 1 Chemical 9 13 5 Stored 24 Renewable 12 Fast Manufactured, mechanical, processed, activated by man Strong/Big/Solid Widely used 14 26 (25)* From coal Used in living things Used by people 13 12 Environmentally clean Used at home 14 17 10 Pollute Immediate From Heat, Explosion or Burning Provides energy Kinetic 17 17 7 Potential I I 14 Stored I I 8 Light 13 10 Sound 8 8 Electrical 8 Performs work 7 8 Power 5 Photosynthesis 7 ~ 8 Makes heat 25 (24)* 16 Associated heat 12 16 12 Associated movement 11 11 Associated electricity 4 7 Associated respiration 6 Associated pollution 12 5 Associated radiation 4 Associated coal 4 Expensive 5 Widely used in Aust 5 11 Used in body or food 17 (I 6) * 10 Used in industry/machines 11 I 0 13 Invisqble 7 9 Easy to harness/converted 12 16 From chemical reactions 5 Not easily stored 7 Not waste product 6 Waste product Efficient 4 For survival 8 4 Useful 4 522 Table 2 (Continued) FETHERSTON Big From plants From earth Controlled Chemical potential energy Other 4 4 4 85 66 68 Note: * Numbers in brackets indicate the number of students holding the construct as some constructs were repeated by students. Table 3 Mean Number of Elicited implementation Constructs, Per Student, From Each Class, Pre- and Post- Rob's class Sean's class Rick's class Statistic Pre Post Pre Post Pre Post (N = 3 I) (N = 29) (N = 27) (N-- 30) (iV = 25) (iV = 29) Mean 4.22 15.59 4.67 14.46 4.76 6.48 Standard dev. 1.41 5.54 1.62 5.11 2.15 2.23 Discussion Accordi ng to Personal Construct Psychol ogy (Kelly, 1955), whi ch views learning as an increase in the number of constructs or in the importance of constructs, the large increase in the number of constructs elicited from each student in Rob and Sean' s classes indicates a similarly large increase in the learning of students in those classes, contrasting markedl y with the learning that has occurred in Ri ck' s class. There is also evidence that school science ideas have been translated into students' personal knowledge in Sean and Rob' s classes. This is provided by the appearance of constructs, whi ch can be classified as science constructs, appearing in students' elicited constructs. For example Potential energy appears in Sean' s and Rob' s class but not in Ri ck' s class. Repert ory grid methodology probes the personal views of students and, as has been inferred from a plethora of studies of misconceptions and students' ideas, students can learn the formal school sci ence without altering their personal ideas very much. The above data supports this contention. However, to further test this idea a school science multiple choice test was administered, before and after implementation, to all three classes and data is reported below. The test was based on the usual test administered to students at the end of the topic dealing with energy and tested the achievement of objectives f or the unit. Comparing all classes bet ween pre- and post-test results showed significant differences for each class pre and post test (F(171, 5~35. 4, p<. 001) . For all three classes there was a significant increase in test scores pre- and post-test but no significant differences bet ween classes. These results indicate that students who learnt using the constructivist approach learnt the school science as well as students who underwent traditional instruction as measured by this multiple choi ce science test. There is no significant difference in the mean scores from the two approaches as measured by a normal school science test. ENERGY AND CONSTRUCTIVIST LEARNING 523 Tabl e 4 Summary Statistics of Students" Performance on Science Test f or Each Class Mean (Pre-test) Std. dev. Mean (Post-test) Std. dev. Rob' s class Sean's class Rick's class (N = 30) (N = 30) (N = 27) 9.3 9.8 8.6 2.6 2.7 2.4 14.5 14.6 13.6 2.3 2.1 2.8 Combining the data from the repertory grid and from the school science test, the conclusion that can be reached is that the two are measures of different kinds of knowledge. It is clear that this multiple choice school test is not capabl e of determining changes in aspects of a student' s knowledge of energy other than the t aught school science. However significant change in knowledge was identified through the use o f the repertory grid. Accordingly it can be stated that, in this study, the standard school multiple choi ce test was an insensitive instrument for determining the breadth and depth of students' knowl edge other than the formal school science. Further, it can be postulated that much of the learning that occurs in constmctivist (and possibly usual) classrooms is not adequately assessed by instruments such as usual school science tests. This point is elaborated using Hilary, a student from Rob' s class, as an example. Hi l ary scored seven on her pretest and 10 on her posttest, out of total marks of 20, which would lead to her being classified by most teachers as a student who is not good at science. Yet, through interviews, it was established that this student could recognise kinetic, potential, sound, heat, chemical, electrical, solar and stored as examples of energies. She could list many uses for these energies and knew that energy coul d be converted from heat to kinetic, coal to kinetic, coal to electricity, chemical to heat and from electricity to light. Constructs elicited by repertory grid post implementation showed that she knew that energy could be stored, was renewable and has the ability to do work. Her knowledge about ener gy was displayed in her explanation about energy in an electric circuit from an Interview-About-Instances - "electricity is passing through the wires and is kinetic energy. . . and the globe gives o f f light and there' s light energy...the globe and there' s stored energy in the battery" - and again displayed in her comments about ice melting - "heat is causing the ice to melt so the water moves to produce kinetic energy." With a representation of a ball rolling, she wrot e "kinetic energy was used to roll the ball, potential is when the ball is going to roll of f the table and sound is when t he ball hits the ground." Clearly this student knew more about energy than was revealed by the multiple choice test. Reasons for this still need investigation beyond the relationship between the language used on the assessment instrument and the student' s language. Hilary could express her own understandings in her own language but may have had t roubl e translating these meanings into formal, school science language. Hilary had a rich understanding of the topic and the question is raised of how best to assess this personal knowledge. This raises a further question related to the equity of current assessment practices and the relationship bet ween these practices and the kind of student who succeeds at a school which adopts these practices. Accordi ng to Personal Construct Psychol ogy, elicited constructs are the constructs that students would use to run their lives from day to day. Students from Rob and Sean' s classes have a larger number of constructs which they can use in their day-to-day lives and therefore have many more ways of vi ewi ng energy. After t he teaching sequence students from Ri ck' s class hold constructs which are not much different in number (and the increase is not statistically significant) from the constructs they held prior to instruction. This means that the method of instruction has 524 FETHERSTON had little effect on their personal ideas about energy and that they have only an insignificant increase in the number of ideas about energy aiter instruction. Yet many would perform very well on a school science test. It can be concluded that the usual method of instruction has had little effect upon their own ideas, on a class basis. The significant increase in the number of constructs in Rob and Sean' s classes is further demonstrated by the number of categories necessary in each class to group the elicited constructs. Constructs from Pi ck' s class could be grouped into 14 groups after instruction but Sean and Rob' s classes required 33 and 34 groups respectively (including the category "Other"). This means that students in Rob and Sean' s classes not only have a larger number of constructs that they can use in relation to energy but also have a more diverse range of constructs upon which to draw. Conclusion In answer to the research questions posed it can be stated that students who experienced the constructivist teaching approach held, on average, more constructs than students taught with usual methods. Students who experienced the constructivist approach incorporated more school science constructs into their construct systems. It can be concluded that students taught with the constructivist approach have increased personal knowledge concerning energy. Because of the larger number of constructs held by students in Rob' s and Sean's classes about energy, they will be more able to encompass future learning events concerning energy within their construct system. In terms of learning, the rich will get richer. Correspondence: Dr Tony Fetherston, School of Education, Edith Cowan University, Mount Lawley Campus, 2 Bradford Street, Mount Lawley, Perth, 6050, Western Australia, Australia. Internet email: t.fetherston@cowan.edu.au References Cosgrove, M., & Osborne, IL J. (1985). Children's science. In R- J. Osborne & P. Freyberg (Eds.), Learning in science: The implications of children's science (pp. 5-14). Auckland: Heinmann. Driver, R. (1989). Students' conceptions and the learning of science. International Journal of Science Education,//(special issue), 481-490. Driver, IL, & Bell, B. (1985). Students' thinking and the learning of science: A constructivist view. School Science Review, 67, 443-456. Fetherston, A. R. (1997). The derivation of a learning approach based on personal construct psychology. International Journal of Science Education, 19(7), 801-819. Fetherston, A. 1L (Submitted). Questions-about-events. Journal of Research in Science Teaching. Fetherstonhaugh, A. R. (1994). Using the repertory grid to probe student's ideas about energy. Research in Science and Technological Education, 12(2), 117-127. Fransella, F., & Bannister, D. (1977). A manual f or repertory grid technique. London: Academic Press. Kelly, G. A. (1955). The psychology ofpersonal constructs (Vols 1 and 2). New York: Norton. Northfield, J., & Symington, D. (1991). Learning in science as personal construction: An Australian perspective. National Key Centre for Science and Mathematics, Curtin University of Technology, Perth. Osborne, R. J., & Gilbert, J. K. (1980). A technique for exploring students' views of the world. Physics Education, 15(6), 376-379. ENERGY AND CONSTRUCTIVIST LEARNING 525 Osborne, R. J., & Freyberg, P. (1985). Learning in science: The implications of children's science. Auckland: Heinemaun. Osborne, R. J., & Wittrock, M. C. (1983). Learning science: A generative process. Science Education, 67, 489-508. Posner, G. J., Strike, K. A., Hewson, P. W., & Gertzog, W. A. (1982). Accomodation of a scientific conception: Toward a theory of conceptual change. Science Education, 66(2), 211-227. Staver, J. K. (1998). Constructivism: Sound theory for explicating the practice of science and science teaching. Journal of Research in Science Teaching, 35(5), 501-520. Watts, M., Jofili, Z., & Bezerra, R. (1997). A case for critical constmctivism and critical thinking in science education. Research in Science Education, 27(2), 309-322.