Argumentation in the Chemistry Laboratory: Inquiryand Confirmatory Experiments

Dvora Katchevich & Avi Hofstein &

Rachel Mamlok-Naaman

Published online: 9 November 2011# Springer Science+Business Media B.V. 2011

Abstract One of the goals of science education is to provide students with the ability toconstruct arguments—reasoning and thinking critically in a scientific context. Over theyears, many studies have been conducted on constructing arguments in science teaching,but only few of them have dealt with studying argumentation in the laboratory. Ourresearch focuses on the process in which students construct arguments in the chemistrylaboratory while conducting various types of experiments. It was found that inquiryexperiments have the potential to serve as an effective platform for formulating arguments,owing to the features of this learning environment. The discourse during inquiry-typeexperiments was found to be rich in arguments, whereas that during confirmatory-typeexperiments was found to be sparse in arguments. The arguments, which were developedduring the discourse of an open inquiry experiment, focus on the hypothesis-building stage,analysis of the results, and drawing appropriate conclusions.

Keywords Argumentation . Chemistry laboratory . Confirmatory-type experiment .

High-order learning skills . Inquiry-type experiment

Theoretical Background

Learning science in a laboratory has a number of features that have contributed toestablishing its centrality in the learning and teaching of science in general and chemistry inparticular (Hodson 1993; Hofstein and Kind in press; Hofstein and Lunetta 2004;Lazarowitz and Tamir 1994; Lunetta 1998; Lunetta et al. 2007). Clearly, the sciencelaboratory, if structured properly, has the potential to develop many important high-orderlearning skills such as asking questions, developing critical thinking, and developingmetacognitive skills. It provides a unique opportunity to collaborate, deliberate, andcommunicate with peers. In a nut shell, it provides an opportunity to learn science by doingscience: hands-on as well as minds-on science.

Res Sci Educ (2013) 43:317–345DOI 10.1007/s11165-011-9267-9

D. Katchevich (*) :A. Hofstein : R. Mamlok-NaamanDepartment of Science Teaching, The Weizmann Institute of Science, Rehovot 76100, Israele-mail: dvora.katchevich@weizmann.ac.il

Over the years, the educational effectiveness of science laboratories as a unique learningenvironment that enables meaningful student learning has been emphasized in manyresearch studies (see, for example, Abrahams and Millar 2008; Hodson 1993; Lazarowitzand Tamir 1994; Lunetta et al. 2007). Moreover, the laboratory provides support for high-order learning skills that include observing, planning an experiment, asking relevantquestions, hypothesizing, and analyzing experimental results (Bybee 2000; Hofstein et al.2004). In this paper, we define science laboratory activities as learning experiences inwhich students interact with materials to observe and better understand the natural world.

Note that assessing the educational effectiveness of the laboratory and its relatedlearning skills requires distinguishing between the different modes of instruction, namely,the nature of the experiments in which the students are involved. Laboratory experimentscan be classified into four types: confirmatory, inquiry, discovery, and conducting anexperiment around a specific problem (in this paper, we will relate solely to the first two).Domin (1999) suggested criteria to define experiments according to the type of resultsobtained from the experiment: the inductive or deductive approach to the activity and,according to who wrote the procedure, either the teacher or the student who must performthe experiment. Other researchers (Fradd et al. 2001; Herron 1971; Schwab 1962)suggested characterizing experiments according to their degree of open-endedness. “Open”in this sense means that the experiment is performed entirely by the student and “closed”means that it is performed entirely by the teacher (e.g. a demonstration). A confirmatoryexperiment is considered “closed” when the students, after learning in the scienceclassroom, perform an experiment that is planned by the teacher. Its approach is deductiveand the results of the experiment are known to both the teacher and students in advance. Incontrast, an inquiry experiment is considered “open” when the students plan how it will becarried out. Its approach is inductive and the results are not known in advance to thestudents and sometimes to the teacher.

Argumentation in the Context of Learning Science

One of the goals of science education is to provide students with the ability to formulatearguments—reasoning and critiquing in a scientific context. Progress in science is partiallybased on arguments and their related rebuttal. Formulating arguments is a particular genre ofdiscourse in which a central epistemological framework is formed as a result of scientificactions. Upon examining the type of activities, it was found that formulating arguments iscentral and significant in developing and conducting science activities. Consequently, it isreasonable to assume that imparting the meaning of scientific content and the essence ofdeveloping a scientific concept would be a way to formulate arguments (Erduran et al. 2004;Hofstein and Kind in press; Hofstein et al. 2008). Scientific language is based on arguments;therefore, students should be provided with opportunities to “talk science” (Lemke 1990). Webelieve that argumentation in a scientific context should be an integral part of this process. Ina classical science lesson teachers ask questions, expect certain answers, and immediatelyevaluate the students’ replies (Cazden 2001). In contrast, working in small groups, in whichthe members are exposed to scientific tasks, provides them with an opportunity to becomeinvolved in a debate and to be supported or rejected by their arguments. During a groupdebate, sometimes with the teacher’s intervention, the group has an opportunity to constructindividual as well as group knowledge. Formulating knowledge in this manner is an exampleof constructivist socio-cultural knowledge, as described by Vygotsky (1978).

According to Jiménez-Aleixandre (2008), the characteristics of an optimal learningenvironment for constructing arguments that relate to students, teachers, curriculum,

318 Res Sci Educ (2013) 43:317–345

assessment, reflection, and communication are as follows: (1) The students must be activein the learning process; they must assess knowledge, establish their claims, and be criticalof others; (2) the teachers have to adopt to student-centered learning, act as a role modelregarding the way they verify their claims, support the development of understanding thenature of knowledge among students, and adopt learning strategies such as inquiry; (3) thecurriculum should incorporate an authentic problem solving approach, which will requirethe students to learn by inquiry; (4) students and teachers should be skilled in assessingclaims, and assessing the students should go beyond written tests; (5) the students should bereflective about their knowledge and understand how it was acquired, and finally (6) thestudents should have an opportunity to conduct a dialogue in which cooperative learningwill take place. Combining these six elements encourages the implementation of anargumentative, interactive learning environment.

From a cognitive perspective, formulating an argument is a conceptual process that canaid in developing an understanding of these concepts. Furthermore, the skill of reasoning,which requires creating a link between claims and evidence, is developed (Osborne 2010).In general, students often have difficulty in formulating arguments; they also have difficultyin selecting and connecting findings that can be used as evidence in supporting their claims(Sandoval and Millwood 2005). Furthermore, students do not formulate high-levelarguments on their own. It is therefore necessary to initiate activities that encourage andsupport formulating arguments, especially with controversial activities that have diversetypes of solutions (Andriessen and Schwarz 2009; Duschl and Osborne 2002). Forexample, Osborne et al. (2004) offered a number of strategies to develop argumentationskills, e.g., exposing students to several explanations regarding a particular scientificsubject and dealing with claims that the students may accept or reject. They based theirassessments on appropriate professional criteria and expose students to two opposingtheories that can explain a particular phenomenon. The students should: (1) explain whatevidence supports each of the theories, (2) construct arguments using structured patternsthat include guiding questions, and (3) predict the experiment’s results, based onappropriate arguments, (4) observe the experiment and explain its results (Predict, Observe,Explain), and (5) design an experiment, carry it out, and discuss the results. Bell and Linn(2000) offered a computerized learning environment for integrating knowledge (KIE) thatintegrates argument construction, and enables the students to search for evidence in order toestablish their claims online.

Other researchers suggested using socio-scientific dilemmas, because these dilemmas areambiguous and enable students to practice the process of simultaneously posing claims andcounter claims (Dawson and Venville 2010; Jiménez-Aleixandre et al. 2000; Sadler 2004;Zohar and Nemet 2002). Building an argument has significant social importance forstudents, in addition to their learning scientific concepts and high-order learning skills.While students are engaged in activities in which they are provided with opportunities todevelop argumentative skills, they learn how to conduct a meaningful conversation withpeers. Needless to say, these skills are useful for overcoming life’s challenges and are notused solely in the context of science learning (Jiménez-Aleixandre et al. 2000).

In recent years, several researchers have used Toulmin’s model (Toulmin 1958) in theirstudies. This model includes three basic components: a claim, evidence, and a warrant forformulating grounded and rational arguments (Bell and Linn 2000; Driver et al. 2000;Erduran et al. 2004; Jiménez-Aleixandre et al. 2000; Kind et al. 2010; Sandoval 2003). Theclaim is an assertion whereby the one who suggests it believes it to be true, e.g., aconclusion, an answer to a question or a problem. Evidence is scientific data that supportthe claim. Scientific data consist of information, such as observations and measurements.

Res Sci Educ (2013) 43:317–345 319

The claim should be based on evidences and the warrant justifies the link between thefindings and the claim. A higher level of argumentation includes a theoretical basis orexplanation at an elementary level, namely, it also includes backing. Similarly, a conditional(qualified) argument or counter claim that intended to refute a particular argument. Arebuttal makes a claim about why certain claims are incorrect and uses additional evidenceand reasoning to justify it.

This model was also adopted for the research study presented in this paper. Walton(1996) developed an alternative framework for analyzing an argument, which characterizesarguments in terms of a scheme of 25 common forms of reasoning. Walton’s frameworkputs more emphasis on the content of an argument, but since this was not the main focus ofour study, we chose not to use it. Note, however, that some researchers have expanded theterm “argument” and include in it the entire scope of reasoning. For example, Means andVoss (1996), regarding reasoning, defined argument as a conclusion supported by at leastone reason.

It is assumed that teaching science through inquiry is an effective teaching strategy forteaching and developing the ability to expand argumentation skills (Duschl and Osborne2002; Wilson et al. 2010). It is assumed that an inquiry activity stimulates the students tobetter understand the research process that scientists undergo. Scientists seek answers tounclear phenomena; they try to explain them by collecting evidence and by constructingarguments. The construction of arguments is a sort of discourse that creates anepistemological framework within the scientific process. When considering the type ofactivities in which scientists engage, one realizes that building significant arguments iscentral to the development of science (Hofstein et al. 2008). Therefore, it was reasonable toassume that we would find evidence for argumentation in the laboratory.

Argumentation in the Science Laboratory

Several researchers (e.g., Gott and Duggan 2007; Sampson and Gleim 2009) who focusedon the issue of argumentation suggested that the inquiry-type laboratory in scienceeducation can provide opportunities for students to develop argumentation skills. However,very little research has been conducted with the goal in mind of accepting or rejecting thisassumption.

For example, Tien and Stacy (1996) found that students who participated in guidedinquiry-type laboratories were better at evaluating evidence obtained from their research.Kelly et al. (1998) analyzed the discourse in a physics laboratory and found that claimsaccompanied by justifications are generally given in response to the claims of a colleague inlight of the experiment’s findings or of the instructions, which may require an explanationor reasoning on the part of the student.

Kim and Song (2006) analyzed the argumentation during and after open-ended inquiriesin middle schools in Korea. They suggested adding instructions that encourage students tosupport their claims and to act as critics similarly to scientists at a conference. Theinstructions refer to the experiment, the writing of the final report, and to the criticalfeedback that the students needed to write to other groups.

Richmond and Striley (1996) claimed that the development of argumentation skills inthe laboratory depends on the type of group. They presented a study, conducted among 10thgrade students, who performed a series of experiments dealing with the ability to cope withthe disease cholera. The students worked in small groups; the researchers found that theargumentation skills that developed depended on the group leader’s personality. In the

320 Res Sci Educ (2013) 43:317–345

groups that had an inclusive leader, all the group members contributed in developing theargumentation, whereas in the groups that had a persuasive leader, it was the leader whodeveloped the argumentation.

Watson et al. (2004) reported about the low quantity of arguments in classes in which theinquiry activities were conducted. Here the laboratory was followed by class discussions inorder to encourage argumentation in the science classroom. They claimed that students inthese situations related to the laboratory as a means by which procedures that would enableobtaining a result, e.g., writing a report, are performed, but they did not relate to thelaboratory as a medium for discussion and decision-making. However, note that somereports also claimed that the discourse in the confirmatory laboratories in secondaryeducation lacks arguments (Abi-El-Mona and Abd-El-Khalick 2006).

Other researchers (Hohenshell and Hand 2006; Keys et al. 1999) suggested a strategy ofbest practice in the laboratory whose outcome is a written report: Science Writing Heuristic(SWH). The lab reports, which are written in this way, should replace the traditional way inwhich students prepare laboratory reports (usually after performing the laboratoryexperiment). The students receive written guidelines that make connections among thecomponents of the inquiry process: observations, posing questions, data collection, andevidence-based claims. The construction of knowledge and the building of relationships aredone by inquiry questions, which help students establish their claims for the data that theygathered. This strategy enables the students to become more active, especially in classroomgroup discussions. Yoon et al. (2010) elaborate on the optimal conditions and specificationsneeded for classroom discussions using the SWH strategy. They claim that a non-threatening learning environment, where students feel comfortable to express themselves, toaccept criticism, to listen to others, and to observe teachers who serve as models, areoptimal conditions for encouraging discourse.

The Study

The main goal of this research study was to explore the high-school chemistry laboratory asa platform for developing and enhancing argumentation. It follows a series of other studiesconducted in Israel that investigated teaching and learning in inquiry-type chemistrylaboratories. More specifically, it focuses on students’ ability to ask questions whileconducting an inquiry-type chemistry experiment (Hofstein et al. 2005), developingmetacognitive skills (Kipnis and Hofstein 2008), investigating students’ attitudes towardand interest in the chemistry laboratory, and students’ perceptions of the chemistrylaboratory learning environment (Dkeidek et al. 2009; Hofstein et al. 2001).

Research Question

In order to investigate students’ argumentative processes in chemistry laboratories, wedecided to investigate how the skills of constructing arguments are expressed in varioustypes of experiments, namely open-ended inquiry and confirmatory-type experiments.

Research Design

The research design presented will refer to some reasons that led to selecting it, namely,implementing a pilot study, selecting a research population and its related characteristics aswell as the research procedures.

Res Sci Educ (2013) 43:317–345 321

The research method is mainly based on the use of qualitative tools. Some of thequalitative findings were analyzed quantitatively. The qualitative approach enabled us todescribe in detail the phenomena and processes that occurred in the laboratory and that arerelated to constructing arguments. Quantitative analysis of the qualitative findings enabledus to describe the magnitude of the phenomena that we identified, with the goal in mind ofcomparing the different types of experiments, namely, the open-ended inquiry experimentvs. the confirmatory one.

Pilot Study

A small-scale study (pilot study) was conducted in one class that consisted of a populationsimilar to the one in the main study. The goal of this study was to investigate whether thediscourse held during the experiment’s procedures initiated argumentation. The discourse,which was audio-taped, was analyzed. Based on the findings, it was clear that the open-ended inquiry experiment served as a potential platform for an argumentative discourse.Therefore, after compiling these initial findings, we planned and conducted the main study.

Research Population of the Main Study

The research population consisted of six classes of 11th and 12th grade chemistry students(N=116) in 5 different high schools in Israel. Note that each class was taught by a differentteacher. The students study in an advanced placement chemistry program that consists of alaboratory unit (about 25% of the total program including students’ final grades in thematriculation examination). All the teachers involved underwent a continuous and intensiveprofessional development program. The laboratory unit lasts two years and includes a seriesof twelve experiments, some of which are open-ended-type inquiry experiments, whereasothers are confirmatory experiments.

In Israel, the chemistry laboratory provides a unique learning environment that differsfrom that in the classroom. In this environment, learning is conducted in small groups (3–4students) in which the students are exposed to various levels and types of laboratoryactivities (Israel Ministry of Education 2007). The different learning skills in which thestudents are involved while conducting the various experiments are detailed in Table 1. Asshown in Table 1, more learning skills are included in the inquiry-type experiment, and theyare more complicated than those used in the confirmatory one.

Laboratory Activities

The experiments in the laboratory include the following: Students perform open-ended-typeinquiry experiments in which they are exposed to a phenomenon; they ask questions aboutit, select the research question, write a hypothesis related to the research question, plan anexperiment in order to examine their hypothesis, and then perform the experiment, organizetheir results, and draw conclusions, as well as analyze and summarize the inquiryexperiment. For more details about the nature of this type of experiment and the type ofactivities in which the students are involved, see Appendix 1.

In contrast, the confirmatory experiments are planned by the teacher with the goal inmind of confirming the theoretical material studied in class. The students perform theexperiments according to the teacher’s instructions, then organize their results, analyzethem, and draw conclusions. For more details about the nature of this type of experimentand the nature of activities in which the students are involved, see Appendix 2.

322 Res Sci Educ (2013) 43:317–345

These activities, especially throughout the open-ended-type inquiry experiments, encouragea discourse that enables the students to develop both cognitive and metacognitive skills (Kipnisand Hofstein 2008). The discourse is based on providing explanations regarding a certainexperimental phenomenon, and as a result, it encourages students to pose arguments.Furthermore (similar to what is done by scientists), at the conclusion stage of eachexperiment, the students must draw conclusions—make claims based on the experimentalfindings, and propose a scientific explanation aligning the findings to the claims.

A confirmatory experiment usually consists of two lessons (45 minutes each), whereasan inquiry experiment usually consists of six lessons. During the first two lessons, studentsare exposed to the phenomenon, ask questions, choose a research question, formulate ahypothesis, and plan an experiment to test the hypothesis. In the next two lessons, thestudents perform the experiment that they designed, collect and analyze the data, drawconclusions, and critically summarize the experiment. At the end of these laboratoryactivities, the students submit a group report. During these four lessons, different kinds ofinteractions between the groups and the teacher take place. Some are initiated by thestudents, who ask for feedback or advice, and others are initiated by the teacher, who isinterested in following up the progress, and in guiding or listening to the group discussion.The groups present their summary to their peers in the last 2 lessons, followed by asummary discussion, in which the teacher emphasizes certain points of the subject matter,as well as the background of the experiment.

Research Tools

The research tools consisted of the following: criterion-based observations in the laboratory,the students’ laboratory reports (“hot reports”), and semi-structured interviews held with thestudents.

Observations in the Laboratory Laboratory observations were conducted during laboratorysessions involving different types of experiments; they focused on the discourse related tothe experiments that took place in the laboratory while students performed the experiments.The discourse was audio-taped and the parts “constructing a rational hypothesis”,“analyzing the results”, and “drawing conclusions” were transcribed. The beginning ofthe various parts was set according to the discourse: students tended to announce when they

Table 1 Skills that are involved during the two types of experiments

Learning skills that are involved during the experiment Confirmatoryexperiment

Open-ended inquiryexperiment

Conducting an experiment according to the teacher’s instructions ✓ ✓

Asking questions ✓

Formulating research questions ✓

Constructing a rational hypothesis ✓

Designing an appropriate inquiry experiment ✓

Conducting the experiment that was planned by the students ✓

Organizing the results ✓ ✓

Analyzing the results ✓ ✓

Drawing conclusions ✓ ✓

Summarizing the experiment’s procedures ✓ ✓

Res Sci Educ (2013) 43:317–345 323

began a particular part. These parts included interactions between the group members, andsometimes interactions between the group members and the teacher, who approached andinteracted with them.

The discourse was analyzed according to the following criteria: the components of the basicargument: claims, evidence, and scientific explanations. The components of the argument wereidentified using Toulmin’s model (Toulmin 1958). Toulmin’s model places more emphasis onthe generic features of the argument, in line with our general interest in argumentation. Inaddition, Toulmin’s model has been used to characterize argumentation in science lessons andis implicit in using the coding system of others (Bell and Linn 2000; Driver et al. 2000;Erduran et al. 2004; Jiménez-Aleixandre et al. 2000; Kuhn et al. 1997; Sandoval 2003).Following these researchers, we adopted Toulmin’s framework to focus on the epistemic andargumentative operations adopted by students. In order to assess the level of the arguments,we chose a tool that refers to the various elements of an argument (see Table 2). This tool waschosen from among many assessment tools appearing in the literature reviewed by Sampsonand Clark’s (2008). This tool is aligned with the discourse style of the laboratory experimentsand with Toulmin’s model; it is based on other tools suggested in former studies (Erduran etal. 2004; Osborne et al. 2004; Simon and Johnson 2008). During the discourse, the studentssuggest different explanations for the various phenomena that they observed during theexperimental procedure and then analyze the data and present arguments. The reliability ofthe coding of the argumentation discourse components was tested in two ways: encoding thecomponents of the argumentation in 20% of the transcribed discourse and checking thereliability using three experts. The percentage of agreement between the experts ranged from85% to 90%. For encoding in which the experts did not agree, judges discussed the issue untilthey reached a consensus. In addition, the authors repeated the encoding; after a while thecorrelation between the early and late coding was 0.95.

The levels of the arguments raised by the students are presented in Table 2. Two majoraspects are referred to: (1) those components that form the basis of the argument (claimevidence and scientific explanations), and (2) the presence of rebuttals or counterclaims.When the argument includes many components, its level is significantly higher. An argumentat level 3 includes the classic elements of an argument: a claim, evidence, and a scientificexplanation that connects them. On the other hand, during an argumentative discourse, thereis an additional dimension that includes a counterclaim or refutation, the presence of whichserves as evidence of a high argumentative discourse level. Consequently, this element wastaken into account when determining argument levels. The highest level of an argument, level5, included a refutation based on accompanying scientific evidence and explanations. Thediscourse analysis was validated by 3 experts. Note that during the analysis of the argument’scomponents, we used a scientific explanation expression instead of a warrant, becausestudents tend to explain the evidence supporting their arguments by using scientificexplanations based on their previous chemistry content knowledge.

The Students’ Laboratory Reports The students’ laboratory reports—“hot reports”—aregroup reports generally written during (or immediately after) the laboratory session(Hofstein et al. 2004). These written reports were collected throughout the year. The reportsections, describing the hypothesis and delineating the rational conclusions, were analyzedin order to determine the components of the arguments and their respective level.

Interviews with the Students A semi-structured interview was conducted at the end of thelaboratory unit, in order to enable us to triangulate the other findings obtained from thelaboratory observations and the students’ laboratory reports. The interviews were conducted

324 Res Sci Educ (2013) 43:317–345

in pairs and included questions about the laboratory unit; they focused on the discoursewhen various types of experiments were performed. A sample of the questions asked andthe interviewed students’ responses will be presented later. The interview was contentvalidated and inter-reliability was performed by 3 experts (researchers) from theDepartment of Science Teaching, the Weizmann Institute of Science. The content validitywas related to questions that were asked, as well as to the categories established by theresearcher (the first author of this paper). The categories were based on the students’answers. The reliability between judges regarding adjusting the classification of students’answers to the categories was done later. The reliability correlation revealed an average of85% among the judges. If there were answers for which the judges did not agree, then theymet and discussed the problems until a consensus was reached.

Table 2 The key to assessing the level of arguments based on (Erduran et al. 2004; Osborne et al. 2004;Simon and Johnson 2008), and some examples

The components Symbol Level Examples of arguments at different levels

Claim C 1 Nurit: The more powder there is the faster the raisins move, andover time [claim].

Claim+Data orClaim+Warrant

CD 2 Nira: The more reactants that there are in the system, the greaterthe concentration of solution B, more products will be obtained,more gas will be generated, more bubbles will be created, andmore raisins will rise [claim+explanation].

CW

Claim+Data+Warrant or Claim+Data+Rebuttal orClaim+Warrant+Rebuttal

CDW 3 Moriah: As we increased the concentration of the solution, therewas a greater amount of sediment [evidence].

Gil: The more we increased the concentration of the solution, themore the quantity of the products increased. We found this byanalyzing the quantity of the solid [claim+evidence].

CDR Moriah: Because the reaction has more reactants, there are morecollisions between the particles of the reactants and consequently,there are more fertile collisions [explanations].

CWR Gil: And then more of the product that forms the solid that weobtained is created and the solution obtained is more turbid[continued explanation combined with evidence].

Claim+Data+Warrant+Backing

CDWB 4 Noam: I want to state that a higher temperature will result in amore frequent occurrence of the reaction [claim]. [He draws agraph] there is an increase in ΔH since this is an endothermicprocess [evidence].

Alon: There is an increase in ΔS as gas is generated; thus, this isa descending graph [evidence+claim].

Noam: At a higher temperature ΔG is more negative and thereaction will be more spontaneous, according to the graph [hepoints to the graph that was drawn in the report].

Alon: The spontaneity will be expressed in a broader dispersion ofthe gas and, as a result, the gas spreads more, because it hasgreater energy.

Ohad: The greater dispersion of the Iodine will be expressed in agreater area that crystallized on the large test-tube [explanation+backing].

Rebuttal thatincludes Claim+Data+Warrant

CDWR 5 Yarden: In the first system, there was no reaction at all [claim]

Bennie: Not so! There was a reaction, but not like in the othersystems. Insufficient gas was generated in order to raise theraisins [refutation based on evidence+explanation].

Res Sci Educ (2013) 43:317–345 325

The Research Procedure

One of the authors of this paper participated in the laboratory lessons, observed them, andrecorded the discourse of the work groups. The teacher determined the work groups, whichwere permanent throughout the research. One or two groups were chosen in each class andtheir discussions were analyzed. The group selections were based on verbalism—groups inwhich a productive discourse took place. These groups were analyzed through differenttypes of experiments: the open-ended inquiry experiments and the confirmatory ones. Theanalyzed laboratory reports belong to the same groups, whose discourses were analyzed aswell. Note that the teachers who participated in this study selected the experiments and theirsequence by themselves and according to the inspection requirements of the laboratory unit.The sequence differed from one teacher to another, but all the teachers distributed theconfirmatory experiments among the inquiry ones.

The results of the classroom laboratory observations and their related analyses reflected5 open-ended inquiry-type experiments and 6 confirmatory-type experiments. In total, 13observation scenarios were collected (part of the experiments were conducted in more thanone class). The classification of the experiments into 2 groups was based on skills specifiedin Table 1. Again, this classification was validated by three experts from the Department ofScience Teaching.

Results and Discussion

With the goal in mind of providing answers to the research question, the results and thediscussion are presented as follows:

& An analysis of a series of the various types of experiments (an open-ended experimentversus a confirmatory experiment).

& The results obtained from interviewing the students.& An analysis of the discourse of an open-ended experiment.& An analysis of the discourse of a confirmatory experiment.

As described in the methods section of this paper, the discourses held in different types ofexperiments were analyzed. The students’ responses (audio recorded) in the variousargumentative parts were transcribed and encoded according to Toulmin’s model (exampleswill be presented later). Thereafter, the transcribed data were divided into episodes. Eachepisode was ranked according to its level of argument (see Table 2). The results of this analysiswere accumulated in order to compare the number of arguments and their related levelresulting from open-ended inquiry-type and confirmatory-type experiments (see Table 3).

As shown in Table 3, which integrates data regarding observations in 11 different groupsin experiments of an open-ended experiment and a confirmatory-type experiment, theaverage number of arguments in an open-ended experiment is significantly larger than thatin a confirmatory study (N is the number of observations made). Furthermore, the averageargumentative level is significantly higher (N is the total number of arguments in the 11observations that were conducted).

As shown in Fig. 1, most of the arguments in the confirmatory-type experiment areindicated as level 1. In other words, the students claim it but do not feel the need to establishit. However, in the inquiry-type experiments most of the arguments are at the 2 and 3 levelsspecified in Table 2 (CD/CW/CDW/CWR/CDR). These are arguments that include a claimbased on evidence, or a scientific explanation or a combination of both of them or one that

326 Res Sci Educ (2013) 43:317–345

includes a refutation. Arguments such as these definitely exemplify an argumentativediscourse. In an open-end experiment, about 14% include refutation (CDR/CWR/CDWR),compared with a confirmatory study experiment, which does not include a discourse thatincludes refutation. There was a significant difference between the frequencies of the variouslevels of the arguments in the discourse of the two types of experiments: # 2

4=13.7 (p=0.001).The main contributions to the difference are levels 1 and 3.

In order to neutralize intervening variables (the teacher, group size, and groupcomposition), which can affect the argumentation discourse, we present in Fig. 2 open-ended inquiry experiments as well as confirmatory ones, which were conducted in the sameschool, same class, taught by the same teacher. Experiments 1 and 2 are open-ended inquiryexperiments, whereas experiments 3 and 4 are confirmatory ones. Figure 2 presents thedistribution of the arguments at different levels of the experiments. As shown, the numberof arguments and their level in the open-ended inquiry experiments are higher than in theconfirmatory ones. Since other intervening variables were neutralized (as previouslymentioned), we can conclude that regarding the experiments performed in the present study,the number of arguments and their level depend on the characteristics of the experiments.

From the analysis of the student interviews conducted at the end of the laboratory unit,we also realized that from the students’ viewpoint, open inquiry experiments serve as atrigger for generating discourse. Those interviewed noted that the scope of the discussionsis much larger in the inquiry experiments than in the confirmatory experiments. Thestudents’ discussions can be characterized as argumentative discourse. In other words,discourse that includes differences of opinion, attempting to persuade others using criticalthinking, and establishing claims based on scientific explanations (Glassner and Schwarz2007; Osborne and Chin 2010), as presented in Table 4.

Table 3 The average level and the average number of arguments that appeared during the discourse in theopen-ended and confirmatory experiments

Type of experiment Mean no. of arguments (SD) Mean level of arguments (SD)

Open-ended experiment 6.0 (2.1) (N=11) 2.41 (1.12) (N=66)

Confirmatory experiment 1.9 (1.2) (N=11) 1.48 (0.60) (N=21)

# 21ðpÞ 12.1 (0.001) 13.5 (0.001)

N in the middle column represents the number of observations related to the average number of arguments(describing the number of observations conducted)

N in the right column represents the average number of arguments (describing the total number of argumentsthat arose during 11 observations)

The Frequency of the Arguments by Level During theDiscourse

0.0

10.0

20.0

30.0

40.0

50.0

60.0

level 1 level 2 level 3 level 4 level 5The level of the argument

Fre

qu

ency

(%

) inquiryconfirmatory

Fig. 1 The frequency of thearguments by level during thediscourse (N=66 argumentsrefer to open-ended inquiryexperiments, and N=21arguments refer to confirmatoryexperiments)

Res Sci Educ (2013) 43:317–345 327

Table 4 presents quotes from interviews with students (N=40) at the end of thelaboratory learning unit, with regard to discussions related to the various research activities.We will refer to one question from the interview regarding the nature of the discussions.

Fig. 2 Arguments by level that were raised during the discourse in the open-ended and confirmatoryexperiments, by the same group. Experiments 1 and 2 are inquiry type, and experiments 3 and 4 areconfirmatory experiments

Table 4 A summary of students’ typical answers (quotes) to the question: Can you explain how thediscussions were conducted in the group during the experiment? This was based on the categories and theirfrequency, which emerged from the interviews (N=40)

Categories Frequencypercentage

Sample answer (quotation)

The group membersattempted topersuade eachother.

67.5% Shani: You and I have had differences of opinion on many occasions;you said this and that and I had a different opinion. From my point ofview, everyone said what they wanted and all of us spoke about thedecisions, such as choosing the research question and a scientific basisfor the hypothesis. Thus, if there was no consensus, we attempted topersuade each other. In some cases, I also felt that the scientific basiswas insufficient and then we sought a solution with the teacher’s help.

Critical Thinking 67.5% Ron: … When we faced a problem, we tried to see what was happening;each person raised an idea. If someone thinks that this is correct heagrees; if someone else thinks that this is incorrect, he must explain hisreasoning. If everyone thinks differently, and if we cannot persuadeeach other, we ask the teacher.

MutualContributions tothe Discussion

95% Lior: During the open-ended experiment, there were discussionsespecially because of the necessity to construct a hypothesis that couldbe established scientifically. Because the hypothesis has to be acceptableto all the group members, each member explains his rationale and whyhe/she thinks this is correct, namely, persuading the group memberswhy the hypothesis is correct.

328 Res Sci Educ (2013) 43:317–345

From the students’ answers, we identified three categories that characterize the discourseduring the performance of the open-ended experiments:

1. There were different opinions during the discourse, and group members attempted topersuade each other.

2. The students displayed critical thinking during the discourse.3. The students reported that there was a mutual contribution to the discourse.

The frequencies presented in Table 4 describe the prevalence of specific categories ofstudents’ answers. The students’ answers usually represent more than one category.

As shown in Table 4, 95% of the students reported a mutual contribution to the discourseand to the group product. Almost 70% indicated that the discourse was an argumentative-type discourse that includes a variety of opinions, an attempt to persuade others, and/orcritical thinking.

Note that a few students occasionally reported some kind of discourse that was notcharacterized as argumentative.

Ron: Someone presents a hypothesis, and if one agrees with it there is almost nodiscussion but sometimes when there is no consensus, there is a discussion.Arnon: I must say that our group has smart people, if I may say so and we generallyagree on the content of the hypothesis, and we think together how to define it.Sometimes there are discussions going on, but usually we agree.Levi: Most of the time, the group accepted my opinion without any argument andthat’s not right. I’d rather be with students who can argue with me, and then I canlearn from them.

Distribution of the Various Parts of Arguments during the Experiments

Based on both the analysis of the discourse as well as the interviews held with thestudents (as was shown in the above section), it is clear that during the various steps ofan inquiry-type experiment one observes different levels of arguments. In the followingsection we will present and describe the distribution of the arguments in different partsof the experiments.

Eleven observations conducted in 5 classes and 4 different open-ended inquiry-typeexperiments (in terms of the number of arguments and their level in the various parts of theongoing discourse) were analyzed in order to determine the distribution of argumentation indifferent parts of the experiments (stages in writing the hypothesis, analyzing the results,and writing conclusions). A total of 66 arguments were found, with a distribution of 31.8%in the hypothesis stage and 68.2 in the analysis stage (including drawing conclusions).Overall, the scopes of the analysis phase and the drawing conclusions phase are larger,although in observing a single experiment, one may obtain other results.

Figure 3 shows the distribution of 21 arguments for the hypothesis stage and 45arguments for the analysis and drawing conclusions stages. The main differences can beseen in the distribution of the arguments in Levels 1 and 3. In the analysis phase thearguments were proven better. This was reflected in the use of artifacts that becameevidence. About 14% of the arguments in the hypothesis phase included the use ofevidence, but 53% of the arguments in the data analysis and the drawing conclusions phasewere evidence-based. This finding reinforces our assumption that the experimentsimplemented in this study have the potential to serve as a suitable platform for building

Res Sci Educ (2013) 43:317–345 329

evidence-based arguments, unlike other studies that indicate that students find it difficult toadopt evidence-based argumentation.

Written Arguments

Various levels of arguments were found in the laboratory written reports (“hot reports”) ofthe inquiry-type experiments, both in the written hypothesis and regarding the part in whichthe conclusions were drawn. The arguments written in the reports originated from the groupdiscourse and they, in fact, reflect the group’s knowledge. A breakdown of the argumentsfound in the reports shows that 36.4% are hypotheses and 63.6% are conclusions. Examplesof arguments at various levels from the reports of the experiment, an open-endedinquiry experiment that dealt with the phenomena “The contact between liquids”, arepresented in Table 5. The written arguments do not include any rebuttal components,

Fig. 3 Distribution of the arguments, by levels, in different parts of the discourse during the open-ended inquiryexperiments: 21 arguments for the hypothesis stage and 45 for the analysis and drawing conclusions stages

Table 5 Examples of written arguments from the reports of students of an open-ended-type inquiryexperiment, and their related arguments (component and level)

Argument from the reports Argumentcomponents

Level

Hypothesis: The higher the ethanol/water ratio, the smaller the water’s surface tension[claim]. The ethanol breaks the hydrogen bonds of the water; this proves that it has asmall hydrophobic tail and a hydrophilic part that breaks the bonds and dissolves inthem. The greater the quantity of ethanol, the greater the number of hydrogen bondsthat are broken and the water’s surface tension decreases [scientific explanation].

CW 2

Conclusion: The solubility level falls when the length of the hydrophobic residueincreases [claim]. The residue rejects the water and cannot connect with it. Thehydrophobic residue rejects the water because it has van der Waals bonds, in contrastto the water, which has hydrogen bonds. [Explanation]

CW 2

Conclusion: Acids react with water more rapidly than the other substances [claim]. In theexperiment that we conducted, we used two acids that took the least amount of time tocreate a uniform mixture with water [evidence]. The acid molecules have more contactcenters for creating hydrogen bonds with the water, because of their functional group(COOH) [scientific explanation]. Therefore, it follows that acids are the substances thatreact most rapidly with water in creating a homogeneous mixture [repeat claim].

CDW 3

330 Res Sci Educ (2013) 43:317–345

since the report was written after the stage in which the students agreed upon thehypothesis or the conclusions.

The existence of arguments in the reports, resulting from the confirmatory-typeexperiments, was highly dependent on the type of instructions that were given to thestudents by their teacher. Sometimes the arguments that appeared in the reports wereshallow but only at a basic claim level, for example, in the experiments “Determining theformula of a hydrate”, or “Determining the percentage of sugar in soft drinks”:

The hydrate formula is CuSO4·5H2O [claim].It can be deduced that as the concentration of the sugar solution rises, the solutiondensity rises together with it [claim].

In experiments in which the laboratory sheet also included appropriate questions, thearguments appeared at a high level. An example of this is in the confirmatory experimentduring which the students determined the concentration of chloride ion in different types ofwater. The following instructions appeared as part of the sheet of instructions forperforming the experiment:

1. Do the experiment’s results form the basis for determining an opinion regarding thequality of the water?

2. Draw conclusions from the experiment.

Examples of written arguments from the reports of a confirmatory experiment arepresented in Table 6. These examples represent relatively high-level arguments—arguments written as answers to the questions in the sheet of instructions for performingthe experiment.

The findings in Table 7 integrate data from the written arguments in the student reportsof 11 different groups regarding an open-ended and a confirmatory-type experiment. Nosignificant difference was found regarding the frequency of the arguments in the students’reports at the various levels in both types of experiments: # 2

2 =2.5 (p=NS). However, mostof the arguments in the open inquiry experiments (N=44) are at levels 2–3, whereas mostarguments in the reports on the confirmatory experiments (N=16) were at levels 1–2. It isnoteworthy that in the reports of the confirmatory experiments, approximately 25% of thearguments were written as level 3, namely arguments that include; claim, evidence, andscientific explanations, similarly to the level 3 arguments appearing in the reports of theopen inquiry experiments (29%). One possible explanation for this finding is that the

Table 6 Examples of written arguments from the reports of students of a confirmatory experiment, theirrelated arguments (component and level)

Argument from the reports Argumentcomponents

Level

The results obtained form the basis for determining an opinion regarding the quality of thewater [claim]. During the experiment, we conducted titration studies a number of times forthe same sample. Therefore, one can assume that the results that we obtained were logicalbecause of the repeated results [explanation]. Our investigation focused on finding theionic concentration of chlorine that we discovered… [evidence]. All the tests conductedindicated that the water complies with the Israeli standard, i.e., it is suitable for drinkingbut the ionic concentration of other metals must be examined in order to estimate thequality of the water tested [an additional qualified claim].

CDW 3

The tap water at Kibbutz Givat Brenner is preferable for drinking [claim] because the ionicconcentration of the chlorine in it is smaller than that at Beth Elazari [explanation].

CW 2

Res Sci Educ (2013) 43:317–345 331

questions that appear in the confirmatory experiments’ instruction sheet could inducestudents to write answers regarding the structure of a high-level argument (see Table 6).Furthermore, another possible explanation for this finding can be seen in the process thatthe students undergo as part of a laboratory unit for acquiring skills. At least in some cases,students who had been trained to draw conclusions in open inquiry experiments also applythis skill in confirmatory experiments. Despite the discourse, which contained relativelyfew high-level arguments, when the students write their conclusions, some of them aresimilar to those that they recorded in the open inquiry experiments. In the reports of theopen inquiry experiments the arguments include hypotheses and conclusions, whereas inthe reports of the confirmatory experiments, the arguments only include conclusions(because the hypotheses are not part of the requirements in these experiments).

We found a significant difference between the average number of arguments in thereports of the two types of experiments: # 2

2=12.5 (p=0.001). However, the differencebetween the average argument levels in the reports of the two types of experiments failed toreach a significant level.

In order to obtain a more in-depth perception regarding the source for the differencesobtained in the inquiry-type and confirmatory-type experiments, we decided to present theanalysis of the discourse recorded and analyzed in two experiments: “The contact betweentwo fluids”, representing an inquiry-type experiment, and “Solubility of a solid in water andin nonaqueous solvents”, representing a confirmatory-type experiment. It is suggested thatsuch an analysis can provide us with the opportunity to follow the nature of the tasks and asa result, the discourse held by the students in the various groups.

Analysis of the Discourse of an Open-ended-type inquiry Experiment

The activity, “the contact between fluids” (see Appendix 1), was conducted in an 11thgrade class in a regular academic high school in Israel. The activity, an open-endedinquiry experiment, was conducted in small collaborative groups (3–4 students). In thepre-lab phase the teacher, Orly, clarified the objective of the experiment as part of thetopic that the students learned: “This experiment is intended to summarize the concept ofstructures of molecules and bonding, which was studied in class.” At the beginning of thelesson, the students received equipment, materials, and instructions for performing theexperiment (see Appendix 1).

In the experiment entitled, “the contact between fluids”, the students observed thecontact between two fluids, namely, water and ethanol. They were asked to carefullyobserve any changes that occur. At the beginning of the contact between the fluids,“irregular behaviors” were observed—sharp movements of fluids, whereby each fluidattempted to prevent a mixture of the two. Finally, a drop of soap in the “irregular” region,

Table 7 The percentage frequency of arguments according to their level (N=44 arguments refer to open-ended inquiry experiments, N=16 arguments refer to confirmatory experiments), the average argument level,and the average number of arguments in the reports of 11 observations of open-ended inquiry andconfirmatory experiments

Type of experiment Frequency (%) Mean level (SD) Mean no. (SD)

Level 1 Level 2 Level 3

Open-ended 18.2 52.3 29.5 2.11 (0.69) 4.0 (1.5)

Confirmatory 37.5 37.5 25 1.87 (0.80) 1.5 (1.0)

332 Res Sci Educ (2013) 43:317–345

after a few seconds, resolved the problem and the fluids mixed. Prior to the contact betweenthe fluids, the students observed how water and ethanol spread differently on the surface.

Following their observations, the students were asked: (1) to pose questions regardingthe phenomena, (2) to select the inquiry-type question for further investigation, (3) tosuggest a hypothesis, and (4) to plan an experiment that would investigate the hypothesis.One of the groups proposed studying the following research question: How does theethanol/water relationship affect the water’s surface tension?

All members of the group participated in the group discourse. They completed andexpanded each other’s explanations by establishing evidence that originated in observingthe preliminary experiment. The following will exemplify one of the discourse segments.(The numbers in parentheses show the sequence number during the discourse.)

(1) Alon: Our hypothesis is that the greater the quantity of ethanol, the smaller thesurface tension will be [claim].(3) Arnon: Ethanol has a smaller surface tension than water does, so if we mix it withwater, the mixture will have a lower surface tension [evidence+claim].(6) Nir: The ethanol will create hydrogen bonds with the water [scientificexplanation].(18) Arnon: Then if it dissolves, the mixture obtained has a new surface tension thatis less and we have a basis for a hypothesis [evidence+claim].(19) Alon: We have a new substance inside the water, which dissolves; what does thismean? It also has hydrogen bonds; we will have hydrogen bonds between the waterand the ethanol molecules [scientific explanation(30) Nir [Writes]: The ethanol breaks the water’s hydrogen bonds. This can beproved, because it dissolves, and this occurs because of the small hydrophobic tail,which breaks the bonds and dissolves in them. The greater the quantity of ethanol, thehydrogen bonds will be broken to a greater extent, and the water’s surface tensiondecreases [claim+evidence+scientific explanation].

The level of the argument in this discourse is 3 (CDW), namely, an argument thatincludes a claim, evidence, and a scientific explanation that links the evidence to the claim.The argument is agreed upon by all group members. The observations of the preliminaryexperiment provide evidence for the argument, and the group members define and structurethe scientific explanation that will underlie their claim. The explanations in the argumentsposed integrate the macro level (solubility, surface tension) with the micro level (molecules,bonds). The source of the evidence that the students incorporate is the preceding experimentin which they observed the usual spreading of water and ethanol on the Petri dish.Furthermore, we noted that the students knew the contents and concepts required forestablishing the hypothesis (hydrogen bonds, hydrophilic, hydrophobic, solubility, andsurface tension). They understood the requirements of the task—constructing a rationalhypothesis and therefore, after Alon had presented the hypothesis that included only aclaim: “Our hypothesis is that the greater the quantity of ethanol, the less the surfacetension will be,” Nir said to his colleagues: “Come on, let’s rationalize our hypothesis.”And when his colleague proposed a reason that was unacceptable to him, he criticized himby saying: “This is not a reason; we have to give a reason.”

After writing the hypothesis, the students were required to plan an experiment thatwould investigate the hypothesis. Planning the experiment included preparing solutions thatcontain water and ethanol in various proportions and examining the diameter of the areaobtained after putting 6 drops of the resulting mixture on a glass surface. The studentsassumed that if the surface tension drops as a result of the mixing, the area in which the

Res Sci Educ (2013) 43:317–345 333

drops spread would be larger. To their surprise, the results did not correlate with theirhypothesis. Alon said: It lost volume … It didn’t work, it should’ve spread out more;therefore, we have to do it another way”.

The group discourse revolved around finding an explanation for the unexpected results.The experiment’s results, which constitute the evidence to which the group members relate,was the focus of the discourse and therefore, all the arguments include evidence. Thediscourse included five episodes in which an argument was formulated. The argument levelof four of them is 2—CD. These arguments include claims based on evidence. The primaryargument in the discourse, the argument that caused the students to understand an aspect ofthe experiment’s results, is an argument at level 3, which includes the refuting CDR. Therefutation of the argument relates to an erroneous interpretation of the results obtained. Adescription of the development of the discourse follows:

The teacher, who was not present when the error that the students made in planning theexperiment occurred, advised them to conduct an additional experiment to interpret theirfindings. She said: “Do you remember that I did an experiment in class in which Iexamined the volume of a mixture of water and ethanol?” (14) The students performed theexperiment and discovered that the volume of the mixture was smaller than the volume ofits ingredients. As a result of the experiment, they thought that this was why they obtainedthe unexpected results. Alon said:

The results that we obtained in the experiment did not refute our hypothesis butrather the manner in which we decided to measure it, which was not good (17). Wediscovered that our method of testing the surface tension of the water does not workbecause the diameter was smaller and because the volume was smaller. And weproved this by performing an additional experiment in which we examined the totalvolume of water and ethanol (20).

At this stage Arnon and Alon discussed the interpretation of the results. Alon wasconvinced that the volume loss stemmed from creating the mixture, whereas Arnon tried toconvince him that the volume loss in creating the mixture indeed exists, but that it is notsignificant and cannot explain the results that they obtained.

(21) Arnon: When referring to three drops and not 20 cc, then the volume loss isinsignificant [claim].(22) Alon: But we noticed this; we noted the loss of volume [evidence].(24) Arnon: We had a substantial loss [claim]; our loss was not because of a changein the volume that took place [rebuttal].(25) Alon: It is impossible to argue with the facts regarding what happened. Weadded only water and the volume reached 1 cm. We added water with something else,which resulted in less volume than before [evidence].(26) Arnon: There is no connection [claim]. Drops of just water are higher and dropsof other substances are lower in terms of volume [evidence].

When Arnon was aware that he had not convinced his colleague, he resorted toadditional evidence based on his calculation in order to clarify to Alon that his explanationfor the volume loss was incorrect.

(34) Arnon: But the volume loss is insignificant [claim]. Of 10 and 10 we lost 0.4;how many extra percentage points? [calculative evidence](35) Alon: 0.4 out of 20 is 2%.(36) Arnon: So what is lost in 3 drops? [a question that expresses a claim].

334 Res Sci Educ (2013) 43:317–345

(37) Alon: When there is a loss of volume, it is impossible to relate to the diameter asa measure for the loss of surface tension [claim—conclusion].

Alon began to understand his mistake. From his statements he understood that it wasimpossible to explain the results using the volume lost, since the phenomenon of volume lostdisqualifies the method as a way of determining surface tension. Furthermore, later during thediscourse, the group finally understood that the volumes of a drop of water and a drop of ethanolare not identical. Arnon said: “Our mistake was that we related to the drops as a given volumeand a drop of each substance is different in size” [claim] (53). And Yuval added: “Thechange in the size of the drop was visible and we did not relate to this” [evidence] (54).

The discourse was very productive since the students elucidated things for themselves—why the planning of the experiment as it was suggested was erroneous. However, thediscourse lacked a scientific explanation at the micro level. The group members discoveredthat the volume of a drop taken from different fluids is different. They thought that thisphenomenon was associated with surface tension, but they did not try to explain whysurface tension affects the volume of the drop and whether there are additional factors thatcould affect it. Even when Arnon raised the question, Alon’s answer was based on anobservation—experimental evidence, and it was not based on a scientific explanation:

(62) Arnon: In our experiment it was impossible to measure the diameter, but it waspossible to conclude, in principle, that the change in volume of each drop constitutesevidence of a decrease in the drop’s surface tension. But why?(63) Alon: Try to see a drop of water and a drop of alcohol from the same pipette;can you see the difference?

Other groups that performed the experiment in this class and another 11th grade classconducted productive discourses that included a large number of episodes in which groupor individual arguments developed. On the one hand, the students had knowledge ofcontent suitable for coping with the experiment (the intermolecular bond and its effect onthe solubility of substances). However, on the other hand, they were exposed to anexperiment involving phenomena such as surface tension and surface active agents that theyhad not previously studied in class and which required a group discussion in order toelucidate these phenomena. A breakdown of the quantity and the level is shown in Fig. 4. Itcan be seen that there is no uniform pattern. The number of arguments and their level differ

The Quantity and Level of Arguments in Different Groups

0

1

2

3

4

5

6

group 1 group 2 group 3 group 4

No

. of

Arg

um

ents

level1

level2

level3

level4

level5

Fig. 4 The quantity and level of arguments in the four different groups (with two different teachers) in theexperiment “The contact between fluids”

Res Sci Educ (2013) 43:317–345 335

from one group to another. However, an argumentative discourse was developed in eachgroup. In the four groups, 71% of the arguments include integration of evidence, 52%include the three basic ingredients of an argument (claim, evidence, and scientificexplanations), and 14.3% include refutation at some other level.

Based on a detailed description of the discourse, one can conclude that the inquiryexperiments used in the present study evoked arguments that focus on the hypothesisphrasing stage, an analysis of the results, and drawing appropriate conclusions. Adistinction must be made, however, between arguments at the individual level and thosedeveloped by the group during the discourse. Generally, both the personal and grouparguments are constructed from claims that integrated scientific explanations or/andevidence (CD/CW/CDW see Table 2). The counter claim or rebuttals become part of thearguments when the results obtained do not correlate with the group’s hypothesis, are notunderstood automatically, or when the group’s members hold different views. Note that thediscourse that develops between the group members is highly dependent on the inquiryquestion selected for examination by the group. Sometimes the answer to the researchquestion is so clear that no profound discourse develops between the group members andthis is even more so with an argumentative discourse. Similar findings related to discoursethat lacks arguments in the framework of junior high-school students’ laboratoryassignments were found in a study conducted by Kind et al. (2010). These assignmentswere classified by Chinn and Malhotra (2002) as simple inquiry assignments in which only2 variables are dependent on each other.

For example, the discourse that was conducted while writing the group’s conclusionsthat examined the process of popping corn kernels in order to obtain popcorn investigated:“How does temperature affect the time that the popping begins?”

(10) Tamir: According to the graph, the higher the strength of the flame, the shorterthe popping time of the kernels [evidence].(11) Bar: One can conclude that the higher the strength of the flame, the higher thetemperature of the oil [claim].(12) Tamir: The higher the strength of the flame, the faster the temperature of the oilreaches the final temperature [explanation].(13) Anat: The oil reaches the boiling point faster. And then the time that it takes thefirst kernel to pop is shorter [continued explanation].

Since there was no disagreement among the group members, and everything waslogically clear, the discourse was rather short.

Analysis of the Discourse during a Confirmatory-type Experiment

The findings and analysis in this section relate to the experiment “Solubility in water and in noaqueous solvents,” which was conducted in an 11th grade class at another high school. This isa confirmatory experiment in which the students work according to the teacher’s instructionsand afterward, they have to record and organize the experiment’s results, analyze them, anddraw conclusions. The teacher, Dorit, defined the goal of the experiment as a goal in thecontent field: “The objective of the experiment is to revise the subjects of structure andchemistry bonds, with an emphasis on the intermolecular bonds and solubility. These subjectshave already been studied in class.” At the beginning of the lesson, the students receivedequipment, materials, and instructions for performing the experiment (see Appendix 2).

The experiment is designed in the style of a classic confirmatory experiment, whichis intended to confirm what was studied in class regarding the solubility of substances.

336 Res Sci Educ (2013) 43:317–345

When one of the observations does not correlate with what is expected (as a result of amishap in preparing the substances), the teacher said: “The previous observations thatyou obtained did not correlate with what we studied. Up to three carbon atom alcoholtypes should have completely dissolved in water; therefore, the resulting turbidity doesnot correlate with the theory.”

In the first section of the experiment, the students examined the solubility of varioussubstances in water and Cyclohexane. At this stage they were asked to interpret andexplain, and afterward to draw conclusions. The interpretation of the results stage was veryroutine. Evidence was based on the experiment’s results and on a scientific explanation(correct or incorrect) of the result.

Clear explanations were revealed at the declaration level. The following will serve asexamples of utterances that were heard in class during the teacher’s explanations or whenstudents solved exercises:

(9) Nili: Avner, why doesn’t the iodine dissolve in water?(10) Avner: Because it does not create hydrogen bonds.(23) Nili: Ethanol is dissolved in water because the two substances are molecularand create new hydrogen bonds between them.

There were explanations at a higher level, but they also repeat and confirm what wasstudied in class:

(26) Neta: Does it dissolve in both water and Cyclohexane?(27) Nili: Yes, it also has a hydrophobic part that can create van der Waals bondswith Cyclohexane.(28) Avner: In ethanol there is a hydrophilic part, the OH, and a hydrophobic part.

Despite the fact that the students still had time at the end of the lesson, the stage ofdrawing conclusions in a group discourse did not take place. The students related to theanalysis and the interpretation of the results as conclusions. Only one argument was madeduring the discourse: “So both can create van der Waals bonds” [claim—conclusion] (18).Similarly, no generalization was made during the discourse. The discourse includedevidence based on the results of the experiment as well as scientific explanations byintegrating concepts studied in class. In the other groups, the discourse was at an evenlower level. Confirmatory experiments should not be neglected because they provide anadditional opportunity to learn concepts, and to incorporate visual demonstrations thatcontribute to understanding the concepts.

An analysis of the group discourse during both the open-ended-type inquiry experimentsand the confirmatory-type experiments revealed a difference in the nature and extent of thediscourse. The discourse during the open-ended inquiry experiments used in this study wasfound to be rich in arguments, whereas the one during the confirmatory experiments wasfound to be sparse to the point that it completely lacked arguments. The findings about theconfirmatory experiments are similar to those obtained by Abi-El-Mona and Abd-El-Khalick (2006).

Summary, Limitations and Recommendations

Over the past decade, many studies have been conducted with the goal in mind ofdeveloping students’ ability to construct arguments in the context of science learning, butonly a few dealt with studying argumentation in the laboratory.

Res Sci Educ (2013) 43:317–345 337

No general agreement exists among researchers regarding the laboratory’s ability toserve as an appropriate platform for constructing arguments. Some researchers argue that intheir laboratory activities, students spend much time gathering data, and as a result, theircognitive skills are less emphasized (Abrahams and Millar 2008; Kind et al. 2010; Watsonet al. 2004). This is in contrast to other researchers (Kelly et al. 1998; Kim and Song 2006;Tien and Stacy 1996; Hohenshell and Hand 2006; Richmond and Striley 1996) whoreported that the laboratory has potential to encourage the construction of arguments andserves as a medium where evidence can be evaluated.

In view of our research, which focuses on student argument processes in the laboratory,it appears that some inquiry experiments (such as those used in the present study) have thepotential to serve as a platform for formulating arguments because of the unique features ofthe learning environment: working in small groups, which enables group discourse(Lazarowitz and Tamir 1994).

Note that a learning environment, in which open-ended inquiry experiments areperformed, consists of the characteristics indicated by Jiménez-Aleixandre (2008), namely,encouraging the construction of arguments in which the student is at the center of theprocess. The nature of the assignments in the open-ended inquiry experiments encouragesstudents to construct arguments as individuals or in small groups. The requirement toexplain certain phenomena, to choose a research question, to formulate a hypothesis, and toanalyze and draw conclusions are triggers for the group discourse, which consists ofarguments constructed either by the individual or by the group. The arguments combineevidence from laboratory observations with explanations that are usually based onclassroom lessons or were constructed during the group’s discourse concerning conceptsthey learn beyond the classroom. Moreover, students have time to discuss these concepts,so that the potential of the assignment can indeed be expressed.

The assignment requests are expressed in the students’ written reports. Thereasoning underlying the hypothesis as well as the explained conclusions provides abasis for constructing written arguments. The students’ reports consisted of differentargument levels; however, no significant difference existed between the average levelsof the arguments in the reports that referred to different inquiry levels. One possibleexplanation for this could be that the appropriate questions that appear in theconfirmatory experiments’ instruction sheet could induce students to write answersregarding the structure of a high-level argument. Another possible explanation could liein the process that the students undergo in writing reports. Despite the discourse, whichwas relatively scant in high-level arguments, when the students write their conclusionsin the laboratory reports, some of them draw conclusions similar to those in the openinquiry experiments.

However, the responses to the tasks and their requirements during the experimentand in writing the report on the experiment are not clear cut. We reiterate that thelevel of the inquiry in the experiment has the potential to affect the nature of thediscourse. In other words, the higher the inquiry level, the more open the task is, andthe group discourse is therefore able to integrate more arguments and counterarguments.

In order to eliminate intervening variables such as the teacher and the type ofgroup, we selected one group and compared the students’ argumentation level whenthey conducted open-ended inquiry experiments vs. confirmatory experiments. Basedon the findings, we can conclude that there were differences, which we believe aredue to the different assignments given to the student open-ended inquiry experimentsor confirmatory ones.

338 Res Sci Educ (2013) 43:317–345

In order that the students will enter into a discourse during which they will constructwell-established arguments, they must have knowledge of the content that supports thescientific background of the experiment, but, on the other hand, there must be “something”beyond this knowledge in the experiment. The understanding required for an experimenthas to be in the ZPD (Zone of Proximal Development), so that during the discourse, thegroup can discuss the subject and advance the knowledge and understanding of its members(Vygotsky 1978).

In the interviews conducted with students when the laboratory learning unit ended, theyemphasized that the laboratory contributed to a better understanding of the material taughtin class during the group discourse. Moreover, they claimed that the discussions focusedespecially on choosing a research question, writing a hypothesis, and analyzing the results.By describing the nature of the discourse, the students indicated that they occasionallyconducted an argumentative discourse.

Note (and this is one of its limitations) that this study was carried out in differentclasses, taught by different teachers. Although the socio-economic background of thestudents who participated in the study was similar, the teachers had different styles ofteaching, and different views regarding their role in the laboratory in general andinquiry laboratories in particular. The teacher is a significant intervening factor in theinquiry-type experiments. Indeed, the experiments’ instructions are the same but theactual requirements of each of the teachers, as well as the teacher’s place in the groupdiscourse are different. Another limitation stems from the variety of researchexperiments. The laboratory experiment that is chosen and its timing are interveningvariables in the study. The teachers have autonomy in choosing their own sequence oflaboratory units for two years. Different experiments require a different level ofconstructing arguments, and within a particular experiment the level of argumentsdepends on the research question chosen by the group.

One should be aware of the fact that the duration of an inquiry-type experiment issignificantly longer than a confirmatory one. The inquiry-type experiment is conductedover at least 4 lessons, whereas the confirmatory one is conducted over 2 lessons.Therefore, we checked the discourse during the confirmatory experiment from its beginninguntil its end—over the 2 lessons. Regarding the inquiry-type experiments—we referred tothe discourse only when creating a hypothesis, analyzing the data, or drawing conclusions.

Another clarification relates to the tool selected for assessing the level of the arguments.It focuses on the argument’s components and gives great weight to refutation discourse, butit does not provide an answer related to the complexity of the argument, nor to the timedevoted to the students’ discussions. A follow-up study should combine additionalassessment tools, in order to provide a more complete assessment of the quality of theargument.

More research is needed in order to obtain better insight regarding constructingarguments in the laboratory. Such research should analyze a wider range of experiments inthe context of their inquiry level and in the context of their scientific background. It wouldbe interesting to investigate whether the timing of the experiment in the teacher’s teachingsequence (before or after teaching the subject in class) affects the level of the argumentsraised by the students. In addition, we recommend conducting research in which theresearch population consists of two groups: (1) students who will perform inquiry-typeexperiments, and (2) students who will perform confirmatory experiments, and who did notacquire any inquiry skills. We believe that by following this recommendation, we will beable to better explore the difference between those students who obtained inquiry skills andthose who were not exposed to them at all.

Res Sci Educ (2013) 43:317–345 339

Appendices

Appendix 1—Open-ended Inquiry Experiment

The Contact between Liquids

Note: Protective glasses and gloves must be worn!

General Instructions:

& Read all the instructions well before beginning the experiment.& Check that you have all the necessary equipment and materials at your disposal in order

to conduct the experiment.

Pay strict attention regarding:

& fulfilling the instructions for carrying out stage A precisely& recording as many observations as possible& reporting the observations clearly and in a well-organized manner& participation of all group members in carrying out the various tasks& using correct and precise scientific language throughout the course

Equipment and materials:

A Petri dishAbout 30 ml of colored waterAbout 30 ml ethanol3 Pasteur pipettesA bottle of liquid soap

Stage A: The Pre-inquiry Experiments

1. Drip colored water with a Pasteur pipette into a Petri dish until it will cover about halfthe area of the base of the plate. Be sure that the other regions are dry.

2. Drip Ethanol with a new Pasteur pipette into the dry part of the plate until the twofluids meet.

3. Describe all the observations. If necessary you can add Ethanol.4. Drip a drop of soap solution into the part where the colored water meets the Ethanol.5. Describe what happens

Stage B: The Inquiry Step

I

1. Formulate 5 varied, relevant questions that arose following the observations thatwere made.

& Choose one of the questions that you would like to investigate.& Formulate this question clearly as an inquiry question, and to the extent possible,

as a link between two variables.

340 Res Sci Educ (2013) 43:317–345

& Clearly formulate a hypothesis that relates to the question that you chose toinvestigate.

& Give reasons for your hypothesis, based on correct and relevant scientificknowledge.

2. Plan an experiment that will check the validity of your hypothesis.

& Detail all the steps of the experiment, including the control stage.& List the equipment and materials needed on the equipment request form.& Consult with the teacher and make changes if necessary.& Submit the list of equipment and materials to the laboratory technician.

II

3. Get the teacher’s approval for the proposed experiment.

& Carry out the experiment that you proposed after receiving the teacher’s approval.& Present the observations and the results in an organized form (table, diagram,

graph, etc.)& Analyze and interpret the results.& Draw conclusions as much as possible based on the experimental results and

rationalize them.& Examine the connection between the inquiry question and the conclusions.

4. In the summarizing group discussion

& Express your opinion about all the stages of the inquiry (limitations, precision,etc.).

& To the extent necessary, point out the changes desirable in the inquiry process.& List additional questions that arose following the whole process.& Prepare your group’s summary of the experiment for presentation before the

class.5. In the summarizing class discussion

Relate to our experiment by considering the reports of all the other work groups.

6. Ensure that the report is well organized, aesthetic, and readable.

Enjoy the work!

Appendix 2—Confirmatory Experiment

Solubility in Water and in non Aqueous Solvents

Note: Protective glasses and gloves must be worn!

General Instructions:

Read all the instructions well before beginning the experiment.Check that you have all the necessary equipment and materials at your disposal beforeconducting the experiment

Res Sci Educ (2013) 43:317–345 341

Equipment and materials:

3 test tubesTest tube supportCyclohexane—C6H12(l)

Ethanol—C2H5OH(l)

Distilled waterIodine—I2(s)Copper Sulphate—CuSO4(s)

Pay attention to:

& precisely fulfill the instructions for carrying out stage A& record as many observations as possible& report the observations clearly and in a well-organized manner& have all the group members participate in carrying out the various tasks& use correct and precise scientific language throughout the course

The experimental procedure

1. Fill a test tube with distilled water up to one-third its height.2. Add a few grains of Copper Sulphate—CuSO4(s) and describe what you observe.3. Add to the same test tube Cyclohexane—C6H12(l), in a volume similar to the volume of

the water. Mix and describe what you observe.4. Add a few grains of Iodine—I2(s) and describe what you observe.5. Fill the second test tube with water up to one-third its height.6. Add to the same test tube Ethanol—C2H5OH(l) in a volume similar to the volume of the

water. Describe what you observe.7. Fill the third test tube with Cyclohexane—C6H12(l) up to one-third its height.8. Add to the same test tube Ethanol—C2H5OH(l), in a volume similar to the volume of

the Cyclohexane. Mix and describe what you observe.

Analyze and interpret the results based on correct and relevant scientificknowledge.Draw conclusions as much as possible based on the experimental results and giveyour reasons.

Questions following the experiment

a. What can we learn from the results of the experiment about the solubility ofCyclohexane in water, and the solubility of Copper Sulphate and Iodine in water or inCyclohexane?

b. Write down the equations of the dissolution reactions.c. What can we learn from the results of the experiment about the solubility of Ethanol in

water and Cyclohexane?d. Write down the equations of the dissolution reactions.e. Explain the results of the experiment using concepts belonging to the subject “structure

and bonding”.

The summarizing group discussionExpress your opinion critically about all the results of the experiment (limitations,

precision, etc.).

342 Res Sci Educ (2013) 43:317–345

References

Abi-El-Mona, I., & Abd-El-Khalick, F. (2006). Argumentative discourse in a high school chemistryclassroom. School Science and Mathematics, 106, 349–361.

Abrahams, I., & Millar, R. (2008). Does practical work really work? A study of the effectiveness of practicalwork as a teaching and learning method in school science. International Journal of Science Education,30, 1945–1969.

Andriessen, J., & Schwarz, B. (2009). Argumentative design. In N. Muller-Mirza & A. N. Perret-Clermont(Eds.), Argumentation and education (pp. 145–174). New York: Springer.

Bell, P., & Linn, M. (2000). Scientific arguments as learning artifacts: designing for learning from the Webwith KIE. International Journal of Science Education, 22, 797–817.

Bybee, R. (2000). Teaching science as inquiry. In J. Minstrell & E. H. van Zee (Eds.), Inquiring into inquirylearning and teaching in science (pp. 20–46). Washington, DC: American Association for theAdvancement of Science.

Cazden, C. (2001). Classroom discourse. Portsmouth: Heinemann.Chinn, C. A., & Malhotra, B. A. (2002). Epistemologically authentic inquiry in schools: a theoretical

framework for evaluation inquiry tasks. Science Education, 86, 175–218.Dawson, V. M., & Venville, G. (2010). Teaching strategies for developing students’ argumentation skills

about socio-scientific issues in high school genetics. Research in Science Education, 40, 133–148.Dkeidek, M., Mamlok-Naaman, R., & Hofstein, A. (2009). Inquiring about the inquiry chemistry laboratory

in Arab high schools in Israel: A comparative study. Paper presented at the meeting of the EuropeanScience Education Research Association, Istanbul, Turkey.

Domin, D. S. (1999). A review of laboratory instruction styles. Journal of Chemical Education, 76, 543–547.doi:10.1021/ed076p543.

Driver, R., Newton, P., & Osborne, J. (2000). Establishing the norms of scientific argumentation inclassrooms. Science Education, 84, 287–312.

Duschl, R. A., & Osborne, J. (2002). Supporting and promoting argumentation discourse in scienceeducation. Studies in Science Education, 38, 39–72.

Erduran, S., Simon, S., & Osborne, J. (2004). TAPping into argumentation: developments in the applicationof Toulmin’s Argument Pattern for studying science discourse. Science Education, 88, 915–933.

Fradd, S. H., Lee, O., Sutman, F. X., & Saxton, M. K. (2001). Promoting science literacy with Englishlanguage learners through instructional materials development: a case study. Bilingual Research Journal,25, 479–501.

Glassner, A., & Schwarz, B. (2007). What stands and develops between creative and critical thinking?Argumentation? Thinking Skills and Creativity, 2, 10–18.

Gott, R., & Duggan, S. (2007). A framework for practical work in science and scientific literacy throughargumentation. Research in Science & Technological Education, 25, 271–291.

Herron, M. D. (1971). The nature of scientific enquiry. The School Review, 79, 171–212.Hodson, D. (1993). Re-thinking old ways: toward a more critical approach to practical work in school

science. Studies in Science Education, 22, 85–142.Hofstein, A., & Lunetta, V. N. (2004). The laboratory in science education: foundation for the 21st century.

Science Education, 98, 28–54.Hofstein, A., & Kind, P. (in press). Learning in and from science laboratories. In B. J. Fraser, K. Tobin, & C.

McRobbie (Eds.), Second international handbook of science education. Dordrecht: Springer.Hofstein, A., Levy, N. T., & Shore, R. (2001). Assessment of the learning environment of inquiry type

laboratories in high-school chemistry. Learning Environments Research, 4, 193–207.Hofstein, A., Shore, R., & Kipnis, M. (2004). Providing high school chemistry students with opportunities to

develop learning skills in an inquiry-type laboratory: a case study. International Journal of ScienceEducation, 26, 47–62.

Hofstein, A., Navon, O., Kipnis, M., & Mamlok-Naaman, R. (2005). Developing students’ ability to askmore and better questions resulting from inquiry-type chemistry laboratories. Journal of Research inScience Teaching, 42, 791–806.

Hofstein, A., Kipnis, M., & Kind, P. (2008). Learning in and from science laboratories: Enhancing students’meta-cognition and argumentation skills. In C. L. Petroselli (Ed.), Science education issues anddevelopment (pp. 59–94). New York: Nova Science.

Hohenshell, L. M., & Hand, B. (2006). Writing-to-learn strategies in secondary school cell biology: a mixedmethod study. International Journal of Science Education, 28, 261–289.

Israel Ministry of Education. (2007). New chemistry curriculum—laboratory unit: Teacher’s guide.Jerusalem: Author. in Hebrew.

Res Sci Educ (2013) 43:317–345 343

Jiménez-Aleixandre, M. P. (2008). Designing argumentation learning environments. In S. Erduran & M. P.Jiménez-Aleixandre (Eds.), Argumentation in science education: Perspectives from classroom-basedresearch (pp. 91–115). Dordrecht: Springer.

Jiménez-Aleixandre, M. P., Rodriguez, A. B., & Duschl, R. A. (2000). “Doing the lesson” or “doingscience”: argument in high school genetics. Science Education, 84, 757–792.

Kelly, G. J., Druker, S., & Chen, C. (1998). Students’ reasoning about electricity: combiningperformance assessment with argumentation analysis. International Journal of Science Education,20, 849–871.

Keys, C. W., Hand, B., Prain, V., & Collins, S. (1999). Using the science writing heuristic as a tool forlearning from laboratory investigations in secondary science. Journal of Research in Science Teaching,36, 1065–1084.

Kim, H., & Song, J. (2006). The features of peer argumentation in middle school students’ scientific inquiry.Research in Science Education, 36, 211–233.

Kind, P., Wilson, J., Hofstein, A., & Kind, V. (2010). Stimulating peer argumentation in the school sciencelaboratory: Exploring the effect of laboratory task formats. Paper presented at the meeting of theNational Association for Research in Science Teaching, Philadelphia.

Kipnis, M., & Hofstein, A. (2008). The inquiry laboratory as a source for development of metacognitiveskills. International Journal of Science and Mathematics Education, 6, 601–627.

Kuhn, D., Shaw, V., & Felton, M. (1997). Effects of dyadic interaction on argumentative reasoning.Cognition and Instruction, 15, 287–315.

Lazarowitz, R., & Tamir, P. (1994). Research on using laboratory instruction in science. In D. L. Gabel (Ed.),Handbook of research on science teaching and learning (pp. 94–130). New York: Macmillan.

Lemke, J. (1990). Talking science: Language, learning, and values. Westport: Ablex.Lunetta, V. N. (1998). The school science laboratory: Historical perspectives and context for contemporary

teaching. In B. Fraser & K. Tobin (Eds.), International handbook of science education (pp. 249–262).Dordrecht: Kluwer.

Lunetta, V. N., Hofstein, A., & Clough, M. P. (2007). Learning and teaching in school laboratory: Ananalysis of research, theory and practice. In S. K. Abell & N. G. Lederman (Eds.), Handbook of researchon science education (pp. 393–441). Mahwah: Erlbaum.

Means, M. L., & Voss, J. F. (1996). Who reasons well? Two studies of informed reasoning among children ofdifferent grade, ability, and knowledge levels. Cognition and Instruction, 14, 139–178.

Osborne, J. F. (2010). Arguing to learn in science: the role of collaborative, critical discourse. Science, 328,463–466.

Osborne, J. F., & Chin, C. (2010). The role of discourse in learning science. In K. Littleton & C. Howe(Eds.), Educational dialogues: Understanding and promoting productive interaction (pp. 88–102).London: Routledge.

Osborne, J. F., Erduran, S., & Simon, S. (2004). Enhancing the quality of argument in school science.Journal of Research in Science Teaching, 41, 994–1020.

Richmond, G., & Striley, J. (1996). Making meaning in classrooms: social processes in small-groupdiscourse and scientific knowledge building. Journal of Research in Science Teaching, 33, 839–858.

Sadler, T. D. (2004). Informal reasoning regarding socio-scientific issues: a critical review of research.Journal of Research in Science Teaching, 41, 513–536.

Sampson, V., & Clark, D. B. (2008). Assessment of ways students generate arguments in science education:current perspectives and recommendations for future directions. Science Education, 92, 447–472.

Sampson, V., & Gleim, L. (2009). Argument-driven inquiry to promote the understanding of importantconcepts & practices in biology. The American Biology Teacher, 71, 465–472.

Sandoval, W. A. (2003). Conceptual and epistemic aspects of students’ scientific explanations. The Journalof the Learning Sciences, 12, 5–51.

Sandoval, W. A., & Millwood, K. A. (2005). The quality of students’ use of evidence in written scientificexplanations. Cognition and Instruction, 23, 23–55.

Schwab, J. J. (1962). The teaching of science as inquiry. In J. J. Schwab & P. F. Brandwein (Eds.), Theteaching of science. Cambridge: Harvard University Press.

Simon, S., & Johnson, S. (2008). Professional learning portfolios for argumentation in school science.International Journal of Science Education, 30, 669–688.

Tien, L. T., & Stacy, M. (1996). The effects of instruction on undergraduate students’ inquiry skills. Paperpresented at the meeting of the American Educational Research Association, New York.

Toulmin, S. (1958). The uses of argument. Cambridge: Cambridge University Press.Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge:

Harvard University Press.Walton, D. N. (1996). Argumentation schemes for presumptive reasoning. Mahwah: Lawrence Erlbaum.

344 Res Sci Educ (2013) 43:317–345

Watson, J. R., Swain, J. R. L., & McRobbie, C. (2004). Students’ discussions in practical scientific inquiries.International Journal of Science Education, 26, 25–45.

Wilson, C. D., Taylor, J. A., Kowalski, S. M., & Carlson, J. (2010). The relative effects and equity ofInquiry-based and commonplace science teaching on students’ knowledge, reasoning, and argumenta-tion. Journal of Research in Science Teaching, 47, 276–301.

Yoon, S. Y., Bennett, W., Mendez, C. A., & Hand, B. (2010). Setting up conditions for negotiation in science.Teaching Science, 56, 51–55.

Zohar, A., & Nemet, F. (2002). Fostering students’ knowledge and argumentation skills through dilemmas inhuman genetics. Journal of Research in Science Teaching, 39, 35–62.

Res Sci Educ (2013) 43:317–345 345