LEARNING ELECTRIC CIRCUITS: THE CONTENT OF ELEMENTARY SCHOOL STUDENTS’ WRITTEN ARGUMENTS

This study aimed to investigate the impact of a teaching intervention for electric circuits on the content of elementary school students’ written arguments. Educational material was constructed based on the constructivist approach to learning with the use of science and engineering practices and was implemented with 34 students aged 11 years. A questionnaire that was provided to students before and after the teaching intervention (pre-test, post-test) was used to collect the data. Data analysis was carried out by classifying the sufficiency of the components of the arguments into levels. The analysis of written answers (arguments) was performed with a framework for assessing the content of arguments (appropriateness of their components: claim, evidence, and reasoning). The data analysis showed that the students significantly improved the content of their arguments.


Introduction
In the new framework for science education of the US National Research Council, the need to involve students in argument from evidence is highlighted (NRC, 2012). Although the construction of arguments by the students has been recognized as important, the research that has been conducted on the quality of students' written arguments is limited (McNeill & Krajcik, 2007;Sampson & Walker, 2012;Songer & Gotwals, 2012). In addition, there are no research papers investigating the quality of students' arguments on electric circuits. This paper focuses on the study of the contribution of a teaching intervention for electrical circuits to the content of elementary school students' arguments.

Literature Review
Research has been carried out thoroughly investigating students' conceptions about electric circuits (Glauert, Esme, & Bridget, 2009;Shipstone, 1988). It has found that students have conceptions that are different from the school science knowledge for electric circuits. Furthermore, research studying the impact of teaching interventions for electric circuits has been conducted and the results of these studies showed that some students during the teaching interventions constructed conceptions about the electrical circuits in the direction of school science knowledge (Afra et al., 2009;Chiu & Lin, 2005). Regarding the construction of arguments, it turned out that students suggest claims without justifying them (Jiménez-Aleixandre et al., 2000;Sadler, 2004) or propose evidence insufficient and inappropriate for justifying the claims (Jiménez-Aleixandre et al., 2000;McNeill & Krajcik, 2012). Furthermore, students rarely use reasoning in the arguments they construct (Mastrogiorgaki & Skoumios, 2018;McNeill & Krajcik, 2012;Songer & Gotwals, 2012).
Although students' conceptions about electric circuits have been investigated and research studying the impact of teaching interventions on students' conceptions has been conducted, there are no research papers studying the impact of teaching interventions to the content of students' arguments for electric circuits.

Purpose and Research Questions
The purpose of this study is to investigate the impact of a teaching intervention for electric circuits, which is based on the constructivist approach to learning with the use of science and engineering practices, on the content of written arguments of elementary school students (11 years old). In particular, the present paper aims to answer the following research questions: a) What is the impact of the proposed teaching intervention on the appropriateness of the claims of elementary school students' written arguments? b) What is the impact of the proposed teaching intervention on the appropriateness of the evidence of elementary school students' written arguments? c) What is the impact of the proposed teaching intervention on the appropriateness of the reasoning of elementary school students' written arguments?

Research Process Phases and Participants
A single group pre-test and post-test quasi-experimental design was adopted. The research processes included two stages. In the first stage, the educational material, and a questionnaire, both related to electric circuits, were developed. In the second stage, the educational material was implemented in the students and the questionnaire was completed before and after the teaching intervention (pre-test, post-test). The educational material was implemented in elementary schools for a period of six weeks, including a total of 18 hours. The research was carried out with the participation of 34 elementary school students of Greece, aged 11 years (18 boys and 16 girls). All the students could speak and write in Greek, while before the teaching intervention, the students had never been taught electric circuits.

Educational Material and Teaching Intervention
The educational material about electric circuits was constructed based on the constructivist view of learning with the use of science and engineering practices. It covered five units: electric circuit, electric current, conductors and insulators, connecting lamps in series, connecting lamps in parallel.
The construction of each unit used the learning model 5Ε by Bybee et al. (2006), which incorporates five phases: engagement, exploration, explanation, elaboration, and evaluation.

a. Engagement
In the engagement phase, the students engaged in activities that pointed to highlight their initial conceptions and help them realize the disagreements they had with each other. Through group discussions, students asked the questions they were going to investigate.

b. Exploration
In the exploration phase, the students became familiar with the processes of planning and carrying out investigations: they asked research questions and made research presumptions, they controlled variables (independent variable, dependent variable, control variables), they reported and implemented labwork processes.

c. Explanation
In the explanation phase, the students processed the data and acknowledged tendencies within the data. It was planned that the students would construct arguments (based on the evidence collected from the labwork). The components of an argument (claim, evidence, reasoning) are presented and explained to the students by the teacher. Moreover, the necessity of constructing arguments was discussed, and the students constructed and evaluated arguments (with the help of self-evaluation sheets and under the guidance of the teacher). The components of the arguments that were presented to the students did not include rebuttal because rebuttal is suggested for secondary education students, after the latter have become familiar with the other three components (claim, evidence, reasoning) (Berland & McNeill, 2010).

d. Elaboration
In the elaboration phase, the students handled problems different from those they had initially worked out so that they could examine the extent to which they systematically activate new knowledge in case of new problems. The students became familiar with activities carried out for identifying the components of the argument, and they constructed and evaluated arguments.

e. Evaluation
In the evaluation phase, the students compared the new knowledge (that they constructed) with their initial conceptions to improve self-control and register their cognitive progress.

Data Collection and Analysis
Data collection used written questionnaire. A small number of students (three 11-yearold students), two elementary education teachers and two science education researchers were provided with the initial version of the questionnaire so that the internal validity of the questionnaire could be ensured. The comments of the above were taken into account in the final form of the questionnaire, which included five problems that asked from students' predictions and justifications for issues related to electric circuits (method of connecting the battery with the lamp in a simple electric circuit, conservation of electric current, conductivity of materials, illumination of lamps connected in series, illumination of lamps connected in parallel). Every problem included one question and data related to the question. The students were asked to answer the question and justify their answers. The Appendix includes a problem about the illumination of lamps connected in parallel. The written arguments produced by the students in their attempt to answer the questions that were included in the pre-test and the post-test constituted the research data. They were allotted one hour to complete each of the questionnaires. A total of 170 written arguments were collected from pre-test and 170 written arguments from post-test.
The evaluation of the content of students' arguments required the appropriateness of the components of students' arguments (claim, evidence, reasoning), when the latter are evaluated regarding school science knowledge. Each component of an argument was classified into one of the two levels (Level 1 and Level 2). A component of an argument (claim, evidence, reasoning) is classified into Level 1 as long as it is absent or inappropriate, while it is classified into Level 2 as long as it is appropriate. It should be noted that the evaluation of arguments was restricted to three out of the four components of the arguments, i.e. the claim, the evidence and the reasoning.
Two arguments used by the students are set out below concerning the question included in the Appendix, accompanied by their evaluations of their contents.

Argument 1: "Maybe, their illumination is affected because the current is shared."
Evaluation of argument 1: As for its content, it includes a claim ("Maybe, their illumination is affected") and a piece of evidence ("the current is shared"). More specifically, a claim considered inappropriate is included (Level 1), inappropriate evidence is included (Level 1), while no reasoning is included (Level 1).
Argument 2: "Their illumination is not affected. When there are two lamps, they provide the same light, and if there are three lamps, they also provide the same light." Evaluation of argument 2: As for the content of the argument, it includes a claim ("Their illumination is not affected"), evidence ("When there are two lamps, they provide the same light, and if there are three lamps, they also provide the same light".) More specifically, a claim considered appropriate is included (Level 2), considered appropriate is included (Level 2), while no reasoning is included (Level 1). Students' arguments were evaluated by two researchers that worked independently. Their differences were settled through discussions. The next step after the arguments were analyzed was to create tables presenting the frequencies and the percentages of the levels that refer to the appropriateness of the components of students' arguments in questionnaires that was handed to the students both before and after the teaching intervention (pre-test, post-test). McNemar's test was used for contrasting the levels (Level 1, Level 2) of the components of students' arguments in the pre-test and the post-test. Table 1 presents the frequencies and the percentages of the levels referring to the appropriateness of claims, evidence and reasoning of students' written arguments in the pre-test and the post-test. With regard to the appropriateness of the claims included in students' arguments in the pre-test and the post-test, it emerged that, while in the pre-test most claims were classified into Level 1 (97.1%), in post-test most claims were classified into Level 2 (58.8%).

Results
As for the appropriateness of the evidence included in students' arguments, it was found that although in the pre-test all the evidence was classified into Level 1 (100%), in the post-test the percentage of evidence classified into Level 1 decreased (70.6%), while the percentage of Level 2 increased (29.4%).
As regards the appropriateness of the reasoning included in students' arguments in the pre-test and the post-test, it was found that although in the pre-test all the reasoning was classified into Level 1 (100%), in the post-test, despite the high percentage classified in Level 1 (79.4%), the percentage classified into Level 2 increased (20.6%).
Furthermore, McNemar's test shows that there is a statistically significant correlation between the appropriateness levels of students' claims [χ²(1)=17.0530, p=0.0001], evidence [χ²(1)=8.1000, p=0.0044] and reasoning [χ²(1)=5.1430, p=0.0233] in the pre-test and the post-test. As a result, a significant improvement was found in the appropriateness of students' claims, evidence and reasoning from the pre-test to the posttest.

Discussions and Conclusion
It was found that the students, before the teaching intervention (as shown by the pretest), produced mainly inappropriate arguments with respect to their content. Most of the students did not suggest appropriate claims, evidence and reasoning. The above results are in line with the results of other studies, which have shown that students enter the educational process already possessing a number of formed conceptions about the world they live and these conceptions differ from scientific knowledge (Driver et al., 1985). Moreover, the above results are in line with the results of other studies, which have shown that the quality of the arguments produced by students of different age is low (McNeill & Krajcik, 2007Songer & Gotwals, 2012). During science teaching the students are usually not taught the components of an argument and rarely are they asked to write and evaluate arguments (Driver et al., 2000).
After the implementation of the teaching intervention (as it resulted from the posttest), it was found that the content of students' written arguments was improved. In particular, the students improved their ability to develop appropriate claims, appropriate evidence supporting the claims, and develop appropriate reasoning, through which they linked the evidence with the claims.
The improvement in the content of students' written arguments could be attributed to the educational material used. Through the activities of the educational material, students had the opportunity to become familiarized with the components of an argument (claim, evidence, reasoning), the way these components are connected with each other as well as the way the students can evaluate an argument. These processes can contribute to improving the quality of arguments (Chen et al., 2016;McNeill & Krajcik, 2012). Furthermore, the activities which allowed the students working in groups to express and elaborate on their conceptions, created the necessary conditions for a discussion among the students. The discussion among the students, in which the students were trying to support their claims and convince their peers through evidence and reasoning, helped the students to actively engage into dialogic argumentation and to improve the quality of students' written arguments (González-Howard & McNeill, 2019). The results of the present research are subject to the restrictions of a small sample, which may not be considered representative of the total population of students. An additional restriction is the use of the questionnaire as the only data collection tool.
The present study was exclusively focused on studying the content of students' written arguments without examining their structure. Further research is required, which will study the progress on the structure of students' arguments and will contrast it with the progress on their content. Moreover, the present study was focused on investigating written arguments. It would be interesting to investigate the progress on students' oral arguments and contrast them with their written arguments. Also, this paper was centered on investigating students' arguments before and after the teaching intervention. It is suggested that the quality of students' arguments be studied during teaching so that students' progression can be investigated.