Printer Friendly

Fostering 5th grade students' understanding of science via salience analogical reasoning in on-line and classroom learning environments.

Introduction

Analogy has been widely noted as being powerful for helping learners to make connections between the pre-existing base knowledge and new target (Driver & Bell, 1986) or to facilitate abstraction from their base knowledge to generate new schema (Duit, 1991) or to employ a combination of logical reasoning processes to assist in problem solving or inquiry (Nersessian, 1992). Many studies have shown that the use of analogy as an instructional approach would improve students' scientific concept learning (Podolefsky & Finkelstein, 2007; Zheng, Yang, Garica, & McCaddden, 2008). Analogical reasoning also plays a prominent role in the field of scientific studies (Hofstadter, 2003). Analogical reasoning may even be "the main engine of inventive thinking" providing the ability to "generate new scientific knowledge, scientific model, and theory for interpretation of scientific phenomenon in scientific inquiry (Glynn, 1991; Hofstadter, 2003). The law of gravitation and atom model structure were developed by Newton and Rutherford-Bohr through the use of analogy to construct their new knowledge of scientific evidence (Dagher, 1998). These imply that analogical reasoning can be very powerful for instruction and learning, and particularly that it is a key skill for students to practice reasoning in learning science, develop their understanding of science, solve scientific problems, and accomplish scientific inquiry. These literatures highlight the prevalence of analogy in facilitating students' scientific conceptual leaning and thinking, however there remains a lack of empirical studies to demonstrate how the use of analogy can facilitate students' analogical reasoning ability as time goes on.

Literature

Analogical reasoning

Analogical reasoning essentially consists of encoding, inference, mapping, application and response (Sternberg, 1977). The Sternberg model focuses on conceptual inference in reasoning. Vosniadou (1989) addressed the idea that analogical reasoning was a process of mapping a known source (analog) system to an unknown target system in order to facilitate the knowledge transfer, and further to make the correlation between the two systems. Gentner (1989) decomposed analogical reasoning into five sub-processes which are: access the base system, perform the mapping between base and target, evaluate the match, store the inferences in target, and extract the commonalities. Whereas Gentner's structure mapping approach predominately focuses on carry-over of prepositional structures, the pragmatic approach of Holyoak (1985) also takes into account contextual factors (Wilbers & Duit, 2001). Gentner views analogical reasoning as a comparison of similarities between base and target. Wilbers and Duit (2001) debates that both Gentner and Holyoaks consider propositionally based knowledge as a starting point for analogy use. As opposed to this, Wilbers and Duit (2001) proposed and claims that intuitive schema and mental models, spontaneously generated by the students when first confronted with the target phenomenon, are essential in analogy use. They further suggest that the analogy is a means of constructing (propositionally based) hypotheses on the basis of (image like) mental models and intuitive schemata triggered by the target phenomenon (Wilbers & Duit, 2001). They also claim that the process of analogy construction, which serves as heuristical exploration for the target, draws on a better known analog that provides some "proto-theory" for the yet unexplored target. The researchers agreed with their ideas that analogies are more of a tool to bring about hypotheses instead of proving them. The proof of hypotheses is a matter of empirical testing and beyond the use of analogy (Bunge, 1973).

Schonborn and Anderson (2008) proposed that it is very important to explicitly explain the relationships between analogy and target concepts to ensure that students interpreted the analogy as relational understanding instead of as a literal depiction of reality. They also addressed that both scientific knowledge and analogical reasoning ability are pivotal for determining whether or not an analogy will foster students' understanding (Schonborn & Anderson, 2008). Thus, analogical reasoning can help students to link the relationship between their own knowledge and new knowledge, and it can encourage students to think and repeatedly compare new and old concepts (May, Hammer, & Roy, 2006). Supported by constant thinking and reasoning, the analogical reasoning ability of students will refine and mature continuously.

The mechanism of analogy has been documented for more than a decade by many psychologists (Gentner, 1989; Vosniadou, 1989). Gentner (1989) studied analogy from similarity perspectives, and she considered analogy to have few attributes and many relations compared to metaphor which has few attributes and few relations. Her theory of analogical mapping emphasizes more on relational similarity than on surface similarity. She suggested that relational similarity is more critical for determining whether or not analogical mapping can be successful. Vosniadou (1989) defined analogical reasoning as a process of identifying, matching and transferring the structural information from base to target though the use of deductive, inductive and analogical reasoning processes. She proposed that objects with salient similarity can have either perceptual attributes or relational similarity and be accessed easily by learners. It remains unclear whether salient similarity with more perceptual attributes or with more relational similarity is more effective for students learning science concepts through the use of analogical reasoning. We are interested in exploring the relative effectiveness of relational salience and surface salience analogical reasoning on student learning. Thus, we specifically designed two different analogies, one being relational salience analogy and the other one being surface salience analogy. Relational salience analogy means that the analogy shares greater relational similarity and less surface (perceptual attributes) similarity with the target in comparison to the surface salience analogy. Surface salience analogy means that the analogy shares greater surface similarity (perceptual attributes) and less relational similarity with the target in comparison to the relational salience analogy.

Difficulty of scientific concept construction involving the image formation of eye

The students' difficulties in learning about light and its role in image formation have been reported by many studies (Bendall, Goldberg, & Galili, 1993; Bouwens, 1987; Fetherstonhaugh, 1990). Their alternative perceptions were summarized in the following: (1) we see something just by light (visual rays), not by light being reflected to our eyes (Fetherstonhaugh, 1990; Shapiro, 1989); (2) the images are not caused by the reflection of light (Rice & Feher, 1987); (3) lenses are not necessary to form images (Fetherstonhaugh, 1990); (4) not sure what is happening between the object and our eyes to form an image with the light (Bendall, Goldberg, & Galili, 1993); (5) not sure about the relationships among eyes, object, and light sources during image formation (Selley, 1996); (6) confused about the refraction of convex lens and its image formation (Saxena, 1991).

A possible reason for students finding it so difficult to learn about light, vision, and image formation might be that these concepts are too abstract and they do not get an opportunity to manipulate and reproduce the process through hands-on science. Though there are a few studies about helping students to develop their understanding of these concepts through different strategies (Fetherstonhaugh & Treagust, 1992; Selley, 1996), there is still little published evidence concerning analogical reasoning as a potential approach to reduce the difficulties of learning abstract concepts such as image formation of the eye. It would be of great interest to know whether or not analogical reasoning can facilitate students' learning of image formation of the eye.

On-line science learning

In the past decade, on-line scientific learning activities have caught the attention of many science educators and researchers. Some studies have documented that on-line learning with well-developed science education theory benefits student learning more than conventional classroom learning does (She & Lee, 2008; Barak & Dori, 2005). However, Dantas and Kemm (2008) designed a physiology laboratory-based e-learning course to facilitate students' learning performance; they found there were no significant differences between the new course with e-learning and the conventional classroom learning. Cole and Todd (2003) suggest that merely transforming the materials from traditional paper-based to internet-based, without including advance designs in instruction content, may not produce better science learning performance.

On the other hand, others studies find that students assigned to a conventional course perform better than web-based course students in their fact-based science concepts (Hansen, Barnett, & MaKinster, 2004) and final exam (Wang & Newlin, 2000). Wang and Newlin (2000) further suggest that the reason why disagreement remains on whether online learning is more effective than conventional instruction on student attitudes and achievement, is that most online science learning programs do not include a well-developed educational theory as their design base. In order to resolve this problem, recent studies started to include well-developed educational theory into their on-line learning program. This resulted in better performance (Liao & She, 2009; Raes, Schellens, Wever, & Vanderhoven, 2012). However, there is a lack of studies regarding whether on-line learning outperforms the classroom instruction when they use the same well-developed science education theory such as salient analogical reasoning theory. Thus, we hope to explore whether on-line learning is more effective than classroom instruction when they all design their learning based upon the salience analogical reasoning theory. In addition, we are more interested in exploring whether surface and relational salience analogical reasoning on-line learning would bring the same success in terms of their scientific concept learning and analogical reasoning.

Purpose

One of the aims of this study was to explore whether the use of on-line learning (on-line as compared to classroom) would influence the progression of scientific concept construction and analogical reasoning processes. A second aim was to investigate the way in which different analogical reasoning (surface and relational) contributes to the progression of scientific concept construction and analogical reasoning processes in mixed methods of analyzing test performance and interview results. In order to fulfil the aims, the following conditions were designed: in the first condition, students participated in surface salience analogical reasoning learning without an on-line aspect. The second and third conditions involve the on-line learning environment, but in slightly different ways. In the second condition, students participated in the same surface salience analogical reasoning learning but in an on-line learning environment. In the third condition (on-line learning environment), students took part in slightly different analogical reasoning--relational salience analogical reasoning - also with an on-line learning environment. The on-line learning groups in the second and third condition are referred to as on-line analogical reasoning groups. A third aim was to examine the relationships among scientific conceptions, scientific reasoning and analogical reasoning.

Methodology Participants

This study included 190 5th grade students (100 boys and 90 girls) from six classes of an urban public elementary school. The science achievement of these students was about average for 5th grade level at their school. Science was taught by the same science teacher across these six intact classes. Each class of students were always together for all the courses in the same classroom, and they were mixed in heterogeneous level of achievement. Six intact classes of students were randomly assigned into three groups and each group with two intact classes of students. Before the experiment, the differences of pre-Scientific Conception Test (SCT), pre-Concept Dependent Analogical Reasoning Test (CDART), and pre-Scientific Reasoning Test (SRT) among three groups were examined to make sure there was no statistical significant difference among these three groups from condition 1 to 3. The mean and SD for each condition were provided in the Table 1. The ANOVA test showed that there was no statistically significant difference among three conditions, regardless of the performance of pre-SCT ([F.sub.(2,187)] = 1.57, p = .212), pre-CDART ([F.sub.(2,187)] = 0.47, p = .624), and pre-SRT ([F.sub.(2,187)] = 0.18, p = .839).

Procedure

The same content of image formation was covered for these three conditions. The difference of instruction for these three conditions were: Condition 1 arranged the presentation of surface salience analogy for learning the target concepts in classroom setting (34 boys and 31 girls), Condition 2 arranged the same presentation of surface salience analogy as condition 1 in the on-line learning environment (34 boys and 29 girls), Condition 3 arranged a slightly different analogical reasoning--relational salience analogy--for learning the target concepts in the on-line learning environment (33 boys and 29 girls). The instruction of each condition lasted an hour.

Pre-test of the Scientific Conception Test (SCT), Concept Dependent Analogical Reasoning Test (CD ART), and the Scientific Reasoning Test (SRT) were administered to all students before the instruction in their regular classroom. A post-test and a retention-test of the SCT and the CDART were administered in their regular classroom after one- and seven-weeks instruction, respectively. In addition, 48 students from four classes (12 students from each class) of condition 2 and 3, who participated in the two slightly different on-line analogical reasoning groups, were interviewed and fully recorded by tape recorder before, one-week after, and 7 weeks after learning. Students were selected to interview based on the following basis: four high achievers (two boys and two girls), four middle achievers (two boys and two girls), and four low achievers (two boys and two girls). Five major questions (Appendix 1) were used to interview students' understanding and analogical reasoning involving the image formation of the eye for about fifteen minutes. Students' interview data were further analyzed according to the correct concepts and categories of analogical reasoning in order to determine whether the nature and extent of their scientific concepts and analogical reasoning ability improved as time went on.

Learning materials

Design of on-line Learning Content: Image formation of Eye

The design of the formation of eye learning materials was based upon the analogical reasoning theory of relational and surface salience analogy. A five-person team participated in the development of the image formation of eye learning materials: one science educator, two middle school science teachers, and two science education graduates.

The same content of image formation of eye and eye problems were covered by three groups which are (1) Structure of target (eye) and analogy (camera or transparent plastic eye model); (2) The mechanism for image formation of target (eye) and analogy (camera or transparent plastic eye model); (3) similarity and difference between image formation of the eye and camera/transparent plastic eye model (Figure 1 & Figure 2); (4) eye problem: the mechanism of causing nearsightedness and presbyopia and the way in which the eye can adjust for these problems with the use of analogy (Figure 3 & Figure 4). For condition 1, the science teacher used the transparent plastic eye model as the analogy to teach the image formation of the eye, and students interacted with the real model in the classroom setting. For condition 2, students learned the same analogy (transparent plastic eye model) through the online learning environment individually. For condition 3, students learned the image formation of eye using the camera as analogy in the on-line learning environment individually. In short, the difference between on-line and classroom instruction is that the on-line condition would visualize the image formation of the eye through on-line visualizing animation, and the classroom condition would learn the image formation of the eye through hands-on activities and lectures.

Instruments

Scientific Conception Test (SCT), Concept Dependent Analogical Reasoning Test (CDART), and Scientific Reasoning Test (SRT)

The SCT and CDART were established by the same panel of five evaluators, ensuring that the items were properly constructed and relevant to the image formation of eye learning materials which students received. The modified SRT (Lawson, 1978) for elementary school students was used in the study (She & Lee, 2008). The SCT, CDART, and SRT all are the two-tier multiple-choice diagnostic instruments and each item contains two tiers. Students need to answer both tiers correctly in order to receive one point. There are 20, 13, and 12 items for SCT, CDART, and SRT, respectively. The pre-test, post-test, and retention-test of SCT and CDART are equivalent. The Cronbach a for the pre-, post-, and retention-SCT were 0.69, 0.80, and 0.81, respectively. The Cronbach a for the pre-, post-, and retention-CDART were 0.72, 0.80, and 0.78, respectively. The Cronbach a of SRT was 0.71 which is close to the Lawson's result of Cronbach a 0.78.

The SCT was developed to measure students' image formation of eye related conceptions (Appendix 2), which required students to choose the answer and justify their response in the first and second tier of the question, respectively. These questions required students to use deeper information processing ability mainly involving analysis and synthesis. The CDART was developed to measure the degree of students' analogical reasoning involving image formation of the eye conceptions (Appendix 3), which required students to select the correct scientific conceptions and use their analogical reasoning in the first and second tier of the question, respectively. The SRT was developed to measure students' scientific reasoning abilities (included: deductive reasoning for aspects of conservation, proportional thinking, identification and control of variables, probabilistic thinking, correlative thinking, and hypothetical deductive), which required students to choose the correct answer and use the scientific reasoning abilities to make correct solution in the first and second tier of the question, respectively.

Interview analysis

Students' interview results were transcribed and analyzed using the coding system developed by the researchers. We evaluated the nature and extent of students' scientific conception and analogical reasoning. Students' scientific conceptions were categorized according their correctness. About one fifth of the interviewee's results were checked by the second coder and inter-rater reliability was 0.90.

Three different categories were used to analyze the nature and extent of analogical reasoning, which were relational analogical reasoning, surface analogical reasoning, and relational reasoning. The inter-rater reliability was 0.93. The coding system used for analogical reasoning was as follows:

* Surface analogical reasoning: This focuses on the students' use of object attributes shared between the target and analogy to reason. For instance, "convex (camera) lens are like eye lens" or "the plastic ball (camera) is transparent (oblong shape), but the eye is not" were coded as surface analogical reasoning.

* Relational analogical reasoning: This focuses on the students' use of relation predicates shared between the target and analogy to reason. For instance, "convex (camera) lens are like eye lens; both of them converge the light and form an inverse real image" was coded as relational analogical reasoning.

* Relational reasoning: This focused on students' use of target object relations to reason without the use of analogy as support. The reasoning level is highest among the four of them, because the students have passed the process of analogy exploring, hypothesis testing and proving; they are beyond the use of analogy. For instance, "It would converge when light passes the lens of the eye, thus the image would form in the retina" was coded as relational reasoning.

Results

Multivariate analysis of the Scientific Conception Test (SCT)

The descriptive statistics and repeated measures of ANOVA were conducted to examine any increase from pre-, post-to retention-SCT for each condition (Table 1). This indicates that the increases in pre-, post- and retention-SCT mean scores also reached statistical significance, regardless of condition 1 ([F.sub.(2,128)] = 16.12***, p = .000), condition 2 ([F.sub.(2,122)] = 54.87***, p = .000) and condition 3 ([F.sub.(2,120)] = 55.97***, p = .000). The post hoc test suggests that the post- and retention-SCT were significantly higher than pre-SCT across three conditions.

The first part was to investigate whether the on-line learning environment or the classroom setting would influence the progression of scientific concept more with the use of surface salience analogy. The Levene test of homogeneity was not significant on pre-SCT ([F.sub.(1,125)] = 0.11, p = .743), thus one-factor MANCOVA was conducted to examine the effects of instructional approaches using post- and retention-SCT scores as the dependent measures, and students' pre-SCT scores as the covariate. The results indicated that instructional approaches (Wilk's A = 0.89, [F.sub.(2,123)] = 7.65 , p = .001) reached a statistically significant effect on the performance of post- and retention-SCT (Table 2). Therefore, the following main effect for instructional approach was performed. The main effect was performed to independently examine the effect of the instructional approaches on post- and retention-SCT. This indicated that the effects for instructional approaches were significant on both post-SCT scores ([F.sub.(1,124)] = 13.48***, p = .000) and retention- SCT ([F.sub.(1,124)] = 8.91**, p = .003). The post-hoc analysis for main effect suggests that the condition 2 groups performed significantly better than the condition 1 group (on-line > classroom, p(post) = .000, p(retention) = .003) on their post- and retention-SCT.

Wilks' lambda is most commonly used statistic for overall significance. It considers differences over all the characteristic roots.

Wilks' lambda is used to examine the relationship between instructional approaches and overall performance on SCT. Wilks' lambda ranges from 0 to 1, and the lower the Wilks' lambda the more the given effect contributes to the model.

The second part was to examine the way in which different analogical reasoning (surface and relational salience) online learning contributes to the progression of scientific concepts. The Levene test of homogeneity was not significant on pre-SCT ([F.sub.(1,121)] = 2.97, p = .087), thus one-factor MANCOVA was conducted. However, there was no statistically significant difference between condition 2 and condition 3 in terms of their post- and retention-SCT (Wilk's A = 0.99, [F.sub.(2,119)] = 0.36, p = .700).

Multivariate analysis of Concept Dependent Analogical Reasoning Test (CDART)

The descriptive statistics and repeated measures of ANOVA were conducted to examine any increase from pre-, post-to retention-CDART (Table 1) for each condition. The result showed an increase in pre-, post- to retention-CDART that also reached statistical significance, regardless of condition 1 ([F.sub.(1,126)] = 22.03 , p = .000), condition 2 ([F.sub.(2,122)] = 28.66***, p = .000) and condition 3 ([F.sub.(2,120)] = 20.23***, p = .000). The post hoc test suggests that the post- and retention-CDART were significantly higher than pre-CDART for all conditions.

The first part was to investigate whether the on-line learning environment or the classroom setting would have a greater influence on the progression of scientific analogical reasoning when they both use the same surface salience analogy. The Levene test of homogeneity was not significant on pre-CDART ([F.sub.(1,124)] = 1.52, p = .220), thus onefactor MANCOVA was conducted to examine the effects of instructional approaches using post- and retentionCDART scores as the dependent measures, and students' pre-CDART scores as the covariate. The results indicated that instructional approaches (Wilk's A = 0.90, [F.sub.(2,122)] = 6.60**, p = .002) had a statistically significant effect on their performance of post- and retention-CDART (Table 3). Therefore, the following main effect for instructional approach was performed. The univariate F (one-factor ANCOVA) was performed to examine the effect of instructional approaches on post- and retention-CDART independently. It indicated that the effects for instructional approach were significant on post-CDART scores ([F.sub.(1,123)] = 12.57**, p = .001), and not significant on retention-CDART. The posthoc indicated that the students in condition 2 performed significantly better than did the condition 1 (p(post) = .001) on their post-CDART.

The second part was to examine the way in which different analogical reasoning (surface and relational salience) online learning contributes to the progression of analogical reasoning. The Levene test of homogeneity was not significant on pre-CDART (F(U 121) = 0.89, p = .346), thus one-factor MANCOVA was conducted. However, there was no statistically significant difference between condition 2 and condition 3 in terms of their post- and retention- CDART (Wilk's [LAMBDA] = 0.97, [F.sub.(2,119)] = 2.15, p = .121).

Stepwise regression analysis

This section examines whether students' degree of conceptual construction would be impacted by their analogical reasoning ability or by scientific reasoning ability. Therefore, the stepwise regression method was used to explore whether the pre-CDART or the pre-SRT would be most important for predicting the post-SCT scores, and whether pre-CDART or the pre-SRT would be the most important factor for predicting the retention-SCT scores.

Results indicated that the best predictor for post-SCT scores was the pre-CDART followed by pre-SRT scores. The standardized regression coefficient for pre-CDART and pre-SRT were 0.33 and 0.30. Together, pre-CDART and preSRT accounted for 24.1% of the variance in post-SCT scores and reached a statistical significance level ([R.sup.2] = .24, [F.sub.(2,186)] = 29.53***, p = .000). Results also showed that the best predictor for retention-SCT scores was pre- CDART followed by pre-SRT scores. The standardized regression coefficient for pre-CDART and pre-SRT were 0.38 and 0.30. Together, pre-CDART and pre-SRT accounted for 28.7% of the variance in retention-SCT scores and reached a statistical significance level ([R.sup.2] = .29, [F.sub.(2,186)] = 37.55***, p = .000) (Table 4).

Analysis of interview

The students in condition 2 and 3 were given five interview questions regarding the conceptions of image formation before, one week after and seven weeks after instruction, and the results were transcribed and further analyzed according to the quantity of correct concepts and level of analogical reasoning.

Quantity of correct concepts

The quantity of correct concepts of interview results from before, one week after, and seven weeks after learning were analyzed. One-factor repeated measures of ANOVA were conducted to examine the effects of instructional approaches and any increase from pre-, post- to retention-correct concepts, respectively. The result indicated that instructional approaches did not reach a statistically significant effect on the performance of pre-, post- to retention-correct concepts. However, it indicated that the increases in pre-, post- to retention-mean score of correct concepts reached statistical significance ([F.sub.(2,90)] = 53.62***, p = .000). The Mauchly's test of sphericity was significant at 0.05, so the Huynh-Feldt of F was used. The post hoc test suggests that the mean scores of correct concepts in the post- and retention-interview results were significantly higher than pre-interview results (p = .000, p = .000) (Table 5). Results indicated that students' mean scores of correct concepts significantly progressed from pre- to post-interview and pre- to retention interview after they received the learning program regardless of condition 3 and condition 2 (Figure 5).

Quantity of analogical reasoning

Three categories of analogical reasoning: surface analogical reasoning (SAR), relation analogical reasoning (RAR), and relational reasoning (RR) were used to measure the quantity of analogical reasoning in which students made progress from pre-, to post-, and then to retention. The descriptive statistic of surface analogical reasoning, relational analogical reasoning, and relational reasoning for students who are in condition 2 and 3 are in Figure 6.

One-factor repeated measures of ANOVA were conducted to examine the effects of instructional approaches and any increase from pre-, post- to retention-SAR, RAR, and RR, respectively. The results indicated that instructional approaches did not reach a statistically significant effect on the performance of pre-, post- to retention-SAR, RAR, and RR, respectively. However, results indicate that the increases in pre-, post- to retention-mean score of SAR, RAR, and RR all reached a statistical significance level ([F.sub.(2,90)(SAR)] = 29.20***, p = .000; [F.sub.(2,90)(RAR)] = 54.96***, p = .000; [F.sub.(2,90)(RR)] = 60.35***, p = .000). The Mauchly's test of sphericity was not significant at 0.05, so the sphericity of F was used. The post hoc test suggests that the mean scores of SAR in post- and retention- interview results were significantly higher than pre-interview results (p = .000, p = .000), regardless of condition 2 or condition 3. A similar pattern was found for the dimensions of RR (p = .000, p = .000), regardless of condition 3 or condition 2. For the dimensions of RAR, the mean scores of RAR in post- and retention-interview results were significantly higher than pre-interview results (p = .000, p = .000), and retention-interview results were significantly higher than postinterview results (p = .000), regardless of condition 3 or condition 2 (Table 6).

Discussion and conclusions

This study documented the findings that the surface salience analogical reasoning on-line learning group (condition 2) significantly outperformed the surface salience analogical reasoning classroom learning group (condition 1) on their students' post-test and retention-test of scientific concept construction and post-test of content dependent analogical reasoning (CDRT). The retention-CDRT did not reach a statistically significant difference level which may indicate that condition 2 has greater immediate instead of retained impact on students' analogical reasoning than does condition 1. Previous studies have demonstrated that classroom instruction with a well-developed science learning theory of surface analogical reasoning is better than conventional instruction without including the surface analogical reasoning theory (Chiu & Lin, 2005; Sarantopoulos & Tsaparlis, 2004), which highlighted the idea that surface analogical reasoning is efficient for promoting student learning in a regular classroom setting. This study moves one step beyond with the demonstration that the on-line learning group outperformed the classroom learning group when they both used the same theories of surface salience analogical reasoning on their scientific concept construction and analogical reasoning ability.

The second focus of this study is to examine any differences existing between relational and surface salience in two different on-line analogical reasoning programs in terms of their analogical reasoning ability and scientific concept construction performance. Our mixed method demonstrated that there were no significant differences in the scientific concept construction and analogical reasoning between the relational salience analogical on-line learning group (condition 3) and the surface salience analogical on-line learning group (condition 2). The result did not support the idea that relational similarity is more critical for deciding if analogical mapping can be successful, which was proposed by Gentner (1989). Vosniadou (1989) proposed that objects with salient similarity can have either perceptual attributes or relational similarity and be accessed easily by learners. Our results seem more likely to support Vosniadou's (1989) ideas that the object with salient similarity can have either perceptual attributes or relational similarity and be easily accessed by learners. Our results move one step beyond to claim that both relational salience and surface salience analogy can have the same efficiency for promoting students' scientific concepts construction and analogical reasoning.

In addition, this study also provides further important information regarding the progression of students' scientific concept construction and analogical reasoning. Qualitative and quantitative results showed that both groups made significant progress from pre- to post- and pre- to retention-performance of concept construction and analogical reasoning. It demonstrated that students' scientific concept construction and analogical reasoning can be promoted significantly over time, regardless of employing surface or relational salience analogical reasoning. It sheds light on the idea that students can benefit through the use of analogical reasoning if the properties of analogy are salient to them. It does not necessarily have to be restricted to relation or surface to be successful.

Our stepwise regression showed other interesting results in that the best single predictor for scientific concept construction is analogical reasoning followed by scientific reasoning. Wilbers and Duit (2001) suggested that analogy is a means of constructing (propositionally based) hypotheses on the basis of (image like) mental models and intuitive schemata triggered by the target phenomenon. The researchers agreed with their ideas that analogy is a tool to bring about hypotheses and find proof for that hypothesis. Other researchers also proposed that analogical reasoning can help students to link the relationship between their own knowledge and new knowledge as well as encourage students to think and repeatedly compare the new concepts and the old concepts (May, Hammer, & Roy, 2006). These explanations provided above support our result that analogical reasoning is a better predictor than scientific reasoning for scientific concept construction in the learning of image formation of the eye.

Analogy has been reported as an effective method to reduce the degree of difficulty of acquiring invisible or abstract scientific concepts (She, 2004; Dagher, 1998). Vosniadou (1989) suggested that analogical reasoning is used to map a known source (analog) system to an unknown target system in order to make a relationship between the systems and to further facilitate knowledge transfer. It provides insight that including analogical reasoning into science teaching can efficiently foster students' concept construction. The camera was the most widely used analogy in the majority science textbook to teach optical concepts about the formation of image and structure of the eye (Glynn, Britton, Semrud-Clikeman, & Muth, 1989; Iding, 1997). This study opens a new window that using a transparent plastic eye model is about as effective for helping students acquire the concepts of image formation and structure of the eye as using the camera. Gentner (1989) proposed that learners need to efficiently encode information at working memory and retrieve information from long-term memory to generate a potential candidate analogy. Once the degree of matching candidate analogy with target reached the appropriate level, analogical reasoning is complete (Gentner, 1989). Researchers have suggested that the use of students' familiar analogy would promote students' analogical reasoning ability (May, Hammer, & Roy, 2006). Our study indicated that the transparent plastic eye model is rather familiar with learners.

Moreover, this study indeed supports findings that the use of on-line learning with analogical reasoning theory fosters better scientific concept understanding and analogical reasoning ability than the classroom setting with the use of the same learning theory. One of the advantages of on-line learning is the easy inclusion of multimedia presentation, which provides students with opportunities to learn science through authentic learning (Miller & Miller, 2000). The design of our surface and relational salience analogical on-line learning was to provide students with authentic science learning opportunity through on-line multimedia presentation. Both surface and relational salience analogical reasoning on-line learning enhance students' scientific concept understanding and analogical reasoning ability with the same effectiveness.

References

Barak, M., & Dori, Y. J. (2005). Enhancing undergraduate students' chemistry understanding through project-based learning in an IT environment. Science Education, 89(1), 117-139.

Bendall, S., Goldberg, F., & Galili, I. (1993). Prospective elementary teachers' prior knowledge about light. Journal of Research in Science Teaching, 30(9), 1169-1187.

Bouwens, F. E. A. (1987). Misconceptions among pupils regarding geometrical optics. In J. D. Novak (Ed.), Proceedings of Second International Seminar: Misconceptions and Educational Strategies in Science and Mathematics (Vol. III, pp. 23- 28). Ithaca, NY: Cornell University

Bunge, M. A. (1973). Method, model, and matter. Boston, MA: Reidel Publishing Co.

Chiu, M. H., & Lin, J. W. (2005). Promoting fourth graders' conceptual change of their understanding of electric current via multiple analogies. Journal of Research in Science Teaching, 42(4), 429-464.

Cole, R. S., & Todd, J. B. (2003). Effects of web-based multimedia homework with immediate rich feedback on student learning in general chemistry. Journal of Chemical Education, 80(11),1338-1343.

Dagher, Z. R. (1998). The case for analogies in teaching science for understanding. In J. J. Mintzes, J. H. Wandersee, & J. D. Novak (Eds.), Teaching science for understanding: A human constructivist view (pp. 195-211). San Diego, CA: Academic Press.

Dantas, A. M., & Kemm, R. E. (2008). A blended approach to active learning in a physiology laboratory-based subject facilitated by an e-learning component. Advances in Physiology Education, 32(1), 65-75.

Driver, R., & Bell, B. (1986). Students' thinking and the learning of science: a constructivist view. School Science Review, 67, 443-356.

Duit, R. (1991). On the role of analogies and metaphors in learning science. Science Education, 75(6), 649-672.

Fetherstonhaugh, T. (1990). Misconceptions and light: a curriculum approach. Research in Science Education, 20(1), 105- 113.

Fetherstonhaugh, T., & Treagust, D. F. (1992). Students' understanding of light and its properties: Teaching to engender conceptual change. Science Education. 76(6), 653-672.

Gentner, D. (1989). The mechanisms of analogical learning. In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp.199-241). New York, NY: Cambridge University Press.

Glynn, S. M. (1991). Explaining science concepts: A teaching-with-analogies model. In S. M. Glynn, R. H. Yeany, & B. K. Britton (Eds.), The psychology of learning science (pp. 219-240). Hillsdale, NJ: Lawrence Erlbaum.

Glynn, S. M., Britton, B. K., Semrud-Clikeman, M., & Muth, K. D. (1989). In J. A. Glover, R. R. Ronning, & C. R. Reynolds (Eds.), Analogical reasoning and problem solving in science textbooks, Handbook of creativity (pp.383-398). New York, NY: Plenum Press.

Hansen, J. A., Barnett, M., & MaKinster, J. G. (2004). The impact of three-dimensional computational modeling on student understanding of astronomical concepts: A quantitative analysis. International Journal of Science Education, 26(11), 1365-1378.

Hofstadter, D. (2003, May). Analogy as the central motor of discovery in physics. Paper presented at a Physics Department Colloquium, Ohio State University, Columbus, Ohio.

Holyoak, K. J. (1985). The pragmatics of analogical transfer. The Psychology of Learning and Motivation, 19, 59-87.

Iding, M. K. (1997). How analogies foster learning from science texts. Instructional Science, 25(4), 233-253.

Lawson, A. E. (1978). Development and validation of the classroom test of formal reasoning. Journal of Research in Science Teaching, 15(1), 11-24.

Liao, Y. W., & She, H. C. (2009). Enhancing eight grade students' scientific conceptual change and scientific reasoning through a web-based learning program. Educational Technology & Society, 12(4), 228-240.

May, D. B., Hammer, D., & Roy, P. (2006). Children's analogical reasoning in a third-grade science discussion. Science Education, 90(2), 316-330.

Miller, S. M., & Miller, K. L. (2000). Theoretical and practical constructions in the design of web-based instruction. In B. Abbey (Ed.), Instructional and cognitive impacts of web-based education (pp. 156-177). Hershey: Idea Group Publishing.

Nersessian, N. J. (1992). How do scientists think? Capturing the dynamics of conceptual change in science. In R. N. Giere (Ed.), Cognitive models of science (Vol. XV, pp. 3-44.) Minnesota studies in the philosophy of science. Minneapolis, MN: University of Minnesota press.

Podolefsky, N. S., & Finkelstein, N. D. (2007). Analogical scaffolding and the learning of abstract ideas in physics: An example from electromagnetic waves. Physical Review Special Topics- Physics Education Research, 3(1), 1-12.

Raes, A., Schellens, T., Wever, B. D., & Vanderhoven, E. (2012). Scaffolding information problem solving in web-based collaborative inquiry learning. Computers & Education, 59(1), 82-94.

Rice, K., & Feher, E. (1987). Pinholes and images: Children's conceptions of light and vision I. Science Education, 71(4), 629639.

Sarantopoulos, P., & Tsaparlis, G. (2004). Analogies in chemistry teaching as a means of attainment of cognitive and affective objectives: A longitudinal study in a naturalistic setting, using analogies with a strong social content. Methods and Issues of Teaching and Learning, 5(1), 33-50.

Schonborn, K. J., & Anderson, T. R. (2008). Bridging the educational research-teaching practice gap- Conceptual understanding part 2- Assessing and developing student knowledge. Biochemistry and Molecular Biology Education, 36(5). 372-379.

Selley, N. J. (1996). Towards a phenomenography of light and vision. International Journal of Science Education, 18(7), 837-846.

Shapiro, B. L. (1989). What children bring to light: Giving high status to learners' views and action in science. Science Education, 73(6), 711-733.

She, H. C. (2004). Facilitating changes in ninth grade students' understanding of dissolution and diffusion through DSLM instruction. Research in Science Education, 34(4), 503-525.

She, H. C., & Lee, C. Q. (2008). SCCR digital learning system for scientific conceptual change and scientific reasoning. Computers & Education, 51(2), 724-742.

Sternberg, R. J. (1977). Component processes in analogical reasoning. Psychological Review, 84(4), 353-378.

Vosniadou, S. (1989). Analogical reasoning as a mechanism in knowledge acquisition: A developmental perspective. In S.

Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 413-437). Cambridge: Cambridge University Press.

Wang, A. Y., & Newlin, M. H. (2000). Characteristics of students who enroll and succeed in psychology Web-based classes. Journal of Educational Psychology, 92(1), 137-143.

Wilbers, J., & Duit, R. (2001). On the micro-structure of analogical reasoning: The case of understanding chaotic systems. Research in Science Education- Past, Present, and Future, 10, 205-210.

Zheng, R. Z., Yang, W., Garcia, D., & McCadden, E. P. (2008). Effects of multimedia and schema induced analogical reasoning on science learning. Journal of Computer Assisted Learning, 24, 474-482.

Appendix 1. Interview Questions

(1) Please describe the similarities and differences between the analogy of the plastic ball/camera and the target of eye.

(2) How can we see objects? How does the eye work?

(3) What happens when light passes through the convex lens and the concave lens?

(4) What is nearsightedness? Why is that people with nearsightedness cannot see objects clearly?

(5) What is\ presbyopia? Why is that people with presbyopia cannot see objects clearly?

Appendix 2. Scientific Conception Test (SCT) Example Question

Please choose the correct answer regarding the pathway of light through the eye.

Because

(a) After light passes through eye lens, it would converge inward and appear the image on the retina.

(b) After light passes through eye lens, it would be reflected and the eye could see any image.

(c) After light passes through eye lens, it would go straight forward and there would be no image on the retina.

(d) After light passes through eye lens, it would spread outward and appear as the fuzzy image on the retina.

Appendix 3. Concept Dependent Analogical Reasoning Test (CDART) Example Question

Of the following diagrams, which statement is correct regarding the similar function among plastic ball, camera, and eye?

(a) Medium of plastic ball, film of camera, and retina has a similar function.

(b) Medium of plastic ball, camera lens, and eye lens have a similar function.

(c) Convex lens, camera lens, and retina have a similar function.

(d) Convex lens, camera film, and eye lens have a similar function.

Because

(a) Convex lens, camera lens, and retina all could show image.

(b) Medium of plastic ball, camera lens, and eye lens all could refract light inward

(c) Medium of plastic ball, film of camera, and retina all could show image,

(d) Convex lens, film of camera, and eye lens all could refract light inward.

Ming-Hua Chuang and Hsiao-Ching She *

Institute of Education, National Chiao Tung University, 1001 Ta-Hsueh Rd, Hsin Chu City 300, Taiwan // scotte0809@msn.com // hcshe@mail.nctu.edu.tw

* Corresponding author

(Submitted March 07, 2012; Revised August 07, 2012; Accepted January 21, 2013)

Table 1. Descriptive statistics of the pre-, post-, and
retention of SCT, CDART, and SRT

               N    Pre-test      Post-test      Retention test
                    Mean   S.D.   Mean    S.D.   Mean     S.D.

Scientific Concept Test (SCT)

Condition 1    65   7.38   3.05   8.69    3.05   9.83     4.38
Condition 2    63   6.94   3.12   10.48   4.72   11.35    4.48
Condition 3    62   6.37   3.52   10.56   4.65   11.39    3.89

Concept Dependent Analogical Reasoning Test (CDART)

Condition 1    65   4.77   2.37   5.62    2.93   6.86     2.46
Condition 2    63   4.95   2.78   7.40    3.00   7.53     3.18
Condition 3    62   4.48   2.97   6.23    3.01   6.98     2.80

Scientific Reasoning Test (SRT)

Condition 1    65   2.88   1.56
Condition 2    63   2.92   2.03
Condition 3    62   2.74   1.64

               Retenion test
               f           p

Scientific Concept Test (SCT)

Condition 1    16.12 ***   .000
Condition 2    54.87 ***   .000
Condition 3    55.97 ***   .000

Concept Dependent Analogical
Reasoning Test (CDART)

Condition 1    22.03 ***   .000
Condition 2    28.66 ***   .000
Condition 3    20.23 ***   .000

Scientific Reasoning Test (SRT)

Condition 1
Condition 2
Condition 3

Condition 1: surface salience classroom group, Condition 2: surface
salience on-line group, Condition 3: Relational salience on-line
group.

Table 2. MANCOVA results of SCT for on-line and classroom groups

Source of           Wilk's     Multivariate   Univariate F
variance            [LAMBDA]   F              Post-test   Retention-
                                                          test
Covariates
Pre-test scores     0.59       43.51 ***
                               (.000)
Group membership
Instructional       0.89       7.65 **        13.48***    8.91**
approaches                     (.001)         (.000)      (.003)

Source of           Post-hoc
variance

Covariates
Pre-test scores

Group membership
Instructional       Post:
approaches          Condition 2 >
                    Condition 1
                    (.000)

                    Retention:
                    Condition 2 >
                    Condition 1
                    (.003)

***p < .001, **p < .01, *p < .05.
SCT: scientific conception test.

Condition 1: surface salience classroom group,
Condition 2: surface salience on-line group.

Table 3. MANCOVA results of CDART for on-line and classroom groups

Source of          Wilk's     Multivariate   Univariate F
variance           [LAMBDA]   F              Post-test

Covariates
Pre-test scores    0.75       20.29 ***
                              (.000)
Group membership
Instructional      0.90       6.60 **        12.57**
approaches                    (.002)         (.001)

Source of                       Post-hoc
variance           Retention-
                   test
Covariates
Pre-test scores

Group membership
Instructional      1.51         Post:
approaches         (.222)       Condition 2 >
                                Condition 1
                                (.001)

***p < .001, **p<.01, *p<.05.
CDART: concept dependent analogical reasoning test.
Condition 1: surface salience classroom group,
Condition 2: surface salience on-line group.

Table 4. Stepwise multiple regression among SCT, CDART, and SRT

Significant Predictor Variable    Standardized   t          p
                                  Regression
                                  Coefficients

Post-test of Scientific Concept Test
  Pre-test of Concept Dependent   0.33           4.95 ***   .000
    Analogical Reasoning Test
  Pre-test of Scientific          0.30           4.50 ***   .000
    Reasoning Test
  Multiple R                      0.49
  [R.sup.2]                       0.24

Retention-test of Scientific Concept Test
  Pre-test of Concept Dependent   0.38           5.94 ***   .000
    Analogical Reasoning Test
  Pre-test of Scientific          0.30           4.66 ***   .000
    Reasoning Test
  Multiple R                      0.54
  [R.sup.2]                       0.29

Significant Predictor Variable    Cumulative %
                                  of Variance
                                  Explained

Post-test of Scientific Concept Test
  Pre-test of Concept Dependent   15.8%
    Analogical Reasoning Test
  Pre-test of Scientific          24.1%
    Reasoning Test
  Multiple R
  [R.sup.2]

Retention-test of Scientific Concept Test
  Pre-test of Concept Dependent   20.4%
    Analogical Reasoning Test
  Pre-test of Scientific          28.7%
    Reasoning Test
  Multiple R
  [R.sup.2]

***p < .001, **p < .01, *p < .05.

Table 5. One-factor repeated measures ANOVA results of
students correct concepts (for interview questions)

Source of variance   N    f          p      Post hoc

Instructional        47   0.05       .818
approaches

Time                 47   53.62***   .000   post>pre(.000),
                                            retention>pre(.000)

Instructional        47   0.79       .456
approaches x
Time

***p < .001, **p < .01, *p < .05.

Table 6. One-factor repeated measures ANOVA results of students'
analogical reasoning (for interview questions)

Source of variance   N    f           p      Post hoc

Surface Analogical
Reasoning

Instructional        47   1.43        .238
approaches

Time                 47   29.20 ***   .000   post > pre (.000),
                                             retention > pre .000)

Instructional        47   0.20        .824
approaches x
Time

Relational
Analogical
Reasoning

Instructional        47   1.29        .262
approaches

Time                 47   54.96 ***   .000   post > pre (.000),
                                             retention > pre (.000),
                                             retention > post (.000)

Instructional        47   1.33        .271
approaches x
Time

Relational
Reasoning

Instructional        47   0.20        .658
approaches

Time                 47   60.35 ***   .000   post > pre (.000),
                                             retention > pre (.000)

Instructional        47   1.08        .343
approaches x
Time

***p < .001, **p < .01, *p < .05.
COPYRIGHT 2013 International Forum of Educational Technology & Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Chuang, Ming-Hua; She, Hsiao-Ching
Publication:Educational Technology & Society
Article Type:Report
Geographic Code:9TAIW
Date:Jul 1, 2013
Words:7602
Previous Article:A novel method for learner assessment based on learner annotations.
Next Article:The increasing acceptance of onscreen marking--the 'tablet computer' effect.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters