Research in Science Education (2023) 53:1063–1079 https://doi.org/10.1007/s11165-023-10122-9 When Science Is Taught This Way, Students Become Critical Friends: Setting the Stage for Student Teachers Paul Nnanyereugo Iwuanyanwu1 Accepted: 21 June 2023 / Published online: 8 July 2023 © The Author(s) 2023 Abstract Effective science education draws on many different ways of teaching science. The liter- ature on science education documents some potential benefits of argumentation instruc- tion as a powerful tool for learning science and maintaining wonder and curiosity in the classroom. Unlike expository teaching, which relies on a teacher-driven pedagogy in which students accept the teacher’s authority over any content to be justified a priori, argumen- tation teaching allows students to focus on the importance of high-quality evidence for epistemic knowledge, reasoning, and justification. Using a quasi-experimental design, two study groups of undergraduate student teachers were exposed to two different learning con- ditions, the Exp-group with dialogic argumentation instruction (DAI) and the Ctrl-group with expository instruction. Each group received the same science content twice a week for 12 weeks (2 h per lesson). Pre- and posttests were administered to collect data. One-way MANCOVA with the pretest results as covariates showed that the instructional approaches (Wilk’s Λ = 0.765, p < 0.001) had a significant effect on the tested variables after the inter- vention. A pairwise comparison of performance indices between the two study groups revealed that the exp-group was better able to evaluate alternative solutions and defend arguments for collaborative consensus on unstructured scientific problems. This suggests that dialogic argumentation instruction can be used to help students improve their scientific reasoning, thinking, and argumentation skills, which are required to solve problems involv- ing scientific phenomena. Keywords Dialogic argumentation · Traditional instruction · Students · Physics · Problem- solving Introduction Over the years, science education research has examined various issues related to science learning, including its nature, content, goals, and problems, and has found that some prob- lems that appear to be specific to science education actually reveal broader instructional * Paul Nnanyereugo Iwuanyanwu paul.Iwuanyanwu@nwu.ac.za 1 Faculty of Education, School of Mathematics, Science and Technology Education, (SDL research unit), Northwest University, Potchefstroom, South Africa Vol.:(0123145 63789) 1064 Research in Science Education (2023) 53:1063–1079 issues that affect students’ interest in science, their achievement, and their perceived use- fulness (Eccles & Wigfield, 2002; Ogunniyi, 2022; Toma & Lederman, 2020). Consid- ering this, scholars and science education reformers recommend that science instruction be tailored to encourage students to reflect critically on science (International Council for Science [ICS], 2011), solve real-world problems, make and defend arguments about sci- entific knowledge (Iwuanyanwu, 2020), and gather and evaluate scientific evidence on a variety of topics to gain new knowledge (Osborne, 2019; Tarekegn et al., 2022). According to the social constructivist perspective (Schreiber & Valle, 2013), the process of gather- ing and analyzing scientific evidence to gain new knowledge requires students to engage in a dialogic relationship with their teachers and peers who contribute to the construction of knowledge through reasoning, thinking, and sense-making (Iwuanyanwu, 2022). When participating in a dialogic relationship with peers, students are able to freely express their views, learn to ask questions, identify assumptions, think through problems with peers, gather and weigh evidence to find solutions, or reach a collaborative consensus. The many benefits of engaging students, especially student teachers, in a dialogic rela- tionship that promotes social construction of knowledge can provide them with the skills they need to function as critical friends within the teaching profession as well as the scien- tific community in general. However, research has shown that student teachers and in-ser- vice teachers have difficulty completing these cognitive tasks (Erduran et al., 2016; Ghebru & Ogunniyi, 2017; Iwuanyanwu, 2017). For example, one of the reasons for conducting the current study was that student teachers in physics lectures were becoming increas- ingly perfunctory in formulating, presenting, and defending arguments for the best solution to unstructured problems and in evaluating evidence from opposing sides based on new data. In the literature, physics teachers who have encountered this type of student difficulty report observations that led them to incorporate dialogic argumentation in their lectures to allow their students and preservice teachers to explore possibilities for social negotia- tion and to discuss the uncertainty of multiple solutions to unstructured scientific prob- lems (Etkina et al., 2019; Gürel & Süzük, 2017; Syafril et al., 2021). In light of this, dia- logic argumentation instruction plays a significant role in equipping student teachers with the methodological repertoire needed to resolve scientific and socio-scientific issues both inside and outside the science classroom (Iwuanyanwu & Ogunniyi, 2018). Considering the above useful suggestions, the present study expands on the exploration of dialogic argumentation instruction in physics classroom, focusing on two study groups of stu- dent teachers (experimental and control groups) as they learn to construct arguments, evalu- ate solutions, and justify them when faced with uncertainty. The following research question guided the study’s data gathering and analysis: How do student teachers in expository and DAI-based classes differ in their ability to develop valid problem-solving strategies, formu- late, present, and defend arguments, and provide reasonable solutions to given problems? Review of Literature Recent research has explored how the argumentation framework can be used in physics education to enhance physics instruction and achieve various learning objectives (Erduran & Park, 2023; Syafril et  al., 2021). Argumentation, which involves asserting claims and using evidence to support them, mirrors as closely as possible the enterprise of science learning (Kuhn & Udell, 2007). According to this perspective, the structure of an argu- ment is determined by how evidence, data, reasons, and claims are presented to support the 1 3 Research in Science Education (2023) 53:1063–1079 1065 argument (Toulmin, 2003), which can follow inductive or deductive reasoning from prem- ises to conclusions (Iwuanyanwu, 2019). Consistent with social constructivism, students’ dialogic interactions with teachers and peers have been shown to promote argumentation in physics classrooms (Gürel & Süzük, 2017; Hansson & Leden, 2016). Thus, the pre- sent study fits within the paradigm of dialogic argumentation (Ghebru & Ogunniyi, 2017; Iwuanyanwu, 2022). In science classrooms, the process of dialogic argumentation becomes evident when a student presents a reason for or against an assertion about a phenomenon (Iwuanyanwu, 2017). In such cases, a common goal is to reach consensus between differ- ent perspectives about plausible or acceptable claims (Erduran & Park, 2023). As Ogun- niyi (2022) points out, learning and teaching science through dialogic argumentation (DAI) provide students and teachers with a forum to express themselves, clarify doubts or anoma- lies, better understand scientific phenomena, and possibly revise their viewpoints based on new knowledge about scientific phenomena. In the current study, the focus on DAI aligns with the objective of helping physics student teachers become better at arguing and defending arguments and producing reasoned solutions to science problems, which are essential skills for the science teaching profession (Erduran & Park, 2023). In learning environments where dialogic argumentation has flourished, student teachers’ drive to explore has also evolved (Iwuanyanwu, 2022). In this regard, DAI approach requires teachers to serve as mediators and ask thought-provoking questions to help students understand science (Iwuanyanwu & Ogunniyi, 2020; Koichu et al., 2022). Conceptual Scheme of Dialogic Argumentation Instruction Dialogic argumentation instruction (DAI) focuses on the components of learning that occur in a socially interactive context in which students learn about subject-specific con- cepts, formulate, present, and defend arguments and counterarguments to resolve conten- tious issues. Essentially, DAI includes three types of arguments that reflect the exploration of classroom activities, beginning with the individual argument or self-talk (intra-locutory arguments). Intra-locutory arguments are driven by a student’s self-talk, wonder, curiosity, or passion to understand and solve a particular scientific problem or phenomenon. During this phase, the student may notice something that intrigues her or stimulates her curiosity, leading her to ask questions, which in turn stimulates her inquiring mind. As she moves through this phase, the student puzzles over the task, asks more questions, and hypoth- esizes to create a new mental framework for the phenomenon. By focusing on the phenom- enon more carefully and resolving related problems, the student may gain a better under- standing of it (Iwuanyanwu, 2020). Consequently, the student provides evidence/data to support claims and/or counterclaims, and at best uses such evidence to justify arguments or solutions to given problems (Belland et al., 2011; Iwuanyanwu, 2022). After completing the individual tasks, students move to their assigned small group to work on the next tasks, which are saturated with arguments and require reflective judg- ments. In this case, students enter into dialogues with their peers (inter-locutory arguments) to address issues that arise from their individual tasks or that are part of the group tasks. In this type of argument, dialogic relationships occur within and between subgroups in which students ask questions, critically engage with each other’s arguments and counterargu- ments, solicit responses, and connect meanings to reach a common consensus. In the cur- rent study, this is evident when students analyze different strategies for solving problems by collecting and comparing critical evidence for opposing viewpoints, making arguments and counterarguments using structures such as if/and/then/but/therefore, or considering 1 3 1066 Research in Science Education (2023) 53:1063–1079 facts for which additional evidence/data is needed to establish a common consensus about the uncertainty of solutions to particular problems (Gürel & Süzük, 2017; Iwuanyanwu, 2022; Jonassen, 2011). When dealing with unstructured problems, a plausible or accept- able solution is derived from the sum of all defended reasonable solutions (Geifman & Raban, 2015; Iwuanyanwu, 2020). Moreover, in large group sessions, decisions made at the individual and group levels are mobilized again and typically articulated by group representatives (trans-locutory argu- ments), and the teacher serves as a mediator to facilitate learning with the explicit goal of reaching consensus. From this perspective, it can be said that teaching physics through dia- logic argumentation can provide student teachers in the current study with a better under- standing of physics that goes beyond the presentation of facts, definitions, laws, and prob- lem-solving skills (Erduran & Park, 2023; Iwuanyanwu, 2019). When physics is taught in the context of dialogic argumentation instruction, it provides insight into students’ prior knowledge and views about specific scientific phenomena (Ghebru & Ogunniyi, 2017), including views that may not be consistent with valid scientific knowledge and may influ- ence the way they approach learning physics concepts (Voss, 2006). In addition, some research evidence agrees that learning physics concepts through DAI can improve students’ ability to present and defend arguments and find reasoned solutions to given scientific problems (Erduran & Park, 2023; Gürel & Süzük, 2017; Iwuanyanwu & Ogunniyi, 2020). Furthermore, the importance of solving scientific problems, making and defending judg- ments regarding the problems, and developing reasoned solutions to the problems cannot be overstated. After all, problem solving is a ubiquitous activity and an indispensable skill for students and teachers to acquire (Iwuanyanwu, 2020). According to UNESCO (2020), real-life challenges and recent COVID-19 health problems have led to an urgent need to address the gaps in students’ ability to think for themselves, reason, and solve different types of problems related to STEM education and other socio-scientific contexts. Recent studies by Erduran and Park (2023) and Tarekegn et al. (2022) suggest that educators could use dialogic argumentation instruction to address these identified gaps, which are among the skills students need in the twenty-first century. In light of this, DAI is a very important tool to help student teachers in the current study think independently; challenge the views of their classmates, lecturers, and others with reasoned arguments and counterarguments; and investigate the unresolved issues related to physics phenomena and problems. By doing so, students can build a broad range of knowledge and inquiry skills and gain more experience to be better able to act as critical friends within a larger scientific community. Methods This study followed a pre–posttest control group design with quantitative data collection and analysis as described below. Setting and Samples The faculty of education at a South African university where this study was conducted offers a four-year science education program that prepares students to become science teachers. The program includes science subjects such as chemistry, physics, and biological sciences. Since the study was prompted by a problem identified in physics education class, it focuses on physics instruction. In the first and second years of the physics education 1 3 Research in Science Education (2023) 53:1063–1079 1067 program, student teachers must take an introductory physics module, usually offered in two separate classes. The module includes learning concepts such as dynamics of uniform circular motion, thermodynamics, fluids, forces and motions, waves, and sound. Student teachers were taught specific topics in two separate classes in their first year and mid-year in their second year, during which the study was conducted. The organization of the lec- tures is such that students attend two hours of physics lectures twice a week for 16 weeks. The participants were second-year physics student teachers (n = 79, 37 females, 42 males) enrolled in the above program. Of the two classes, one was inducted as Exp-group (n = 46) and the other as the Ctrl-group (n = 33). Rural, suburban, and urban student teach- ers were represented in each class. They were mostly between 19 and 23 years old (average age = 20; standard deviation = 3.86). Family income ranged from dual income to middle- income. Student teachers who volunteered to participate in the study after it was approved by the ethics committee gave their consent by completing the POPIA consent form. Using POPIA guidelines, all tasks submitted and assessed were treated with privacy, confiden- tiality, and anonymity. The first step of the study was to use a self-developed research instrument to generate baseline data to compare students in the two classes. Details of the instrument review process are provided under instrumentation. The first application of the instrument (pretest) lasted for two hours in week 1. Pretest data suggests that the study groups did not differ significantly (F = 4.82; p = 0.295). Based on this result, the study groups were inducted according to the pretest–posttest control group design. Instrumentation Seventeen different science problems were developed. The tasks required students to develop valid problem-solving strategies, formulate and defend arguments, evaluate the reasonableness of alternative solutions, and defend reasonable solutions, as shown in Table 1. Most tasks do not have unique correct answers/solutions (see, for example, the science problem SP-Q2). The tasks were designed to elicit agreement and/or disagreement about their solutions, even when the tasks are considered solved. The instrument was tested for validity and reliability by two independent science educators who reviewed the tasks. In this regard, the study instrument was appraised for its content level, language appropriate- ness, and conceptual coverage. After several revisions of the instrument, the final version yielded Cohen’s kappa value of 0.78. The Kuder–Richardson 21 reliability coefficient was 0.73. These data indicate that the overall consistency and reliability of the 17 items were satisfactory, reliable, and appropriate (Creswell, 2013). Each of the 17 items required stu- dents to (a) develop valid problem-solving strategies, (b) formulate defensible arguments for their chosen strategies, (c) judge the reasonableness of alternative solutions (if applica- ble) to resolve ambiguities, and (d) defend the reasonable solutions/collaborative consensus reached. Following the existing literature, it was believed that the combination of the four variables (a–d) could allow for the integration of many of the skills and abilities student teachers need to collect and evaluate data/evidence, formulate and defend arguments, and, at their best, use arguments to solve physics problems and communicate scientific knowl- edge (Adams & Wieman, 2015; Belland et al., 2011; Iwuanyanwu, 2020). Procedure In accordance with the experimental pre–posttest control group design, the two study groups were subjected to a series of biweekly physics lectures lasting 2 h per session over 1 3 1068 Research in Science Education (2023) 53:1063–1079 1 3 Table 1 Students’ use of arguments in scientific problem solving Targeted variables Description of DAI activity Proficiency judgement Construct valid problem-solving strategies Students gather or use available data to generate a set of Students show evidence of generated ideas in a network images of the problem context, as well as organize and of mental spaces in order to read, translate, and make analyze known and/or unknown quantities/variables of sense of the complex set of relationships within the given problems from various sources. Identify, define, context that provide background for the problem and represent the problem to be solved diagram- (Iwuanyanwu, 2022) matically (where applicable). Evaluate the potential strategies in light of viable or acceptable alternatives (Asterhan & Swartz, 2007; Iwuanyanwu, 2020) Formulate defensible arguments for chosen strategies Students use data to make claims and/or counterclaims Students used text structures such as if/then/but/then to (I agree/disagree with…), back up claims with build a supporting argument for their strategy selection evidence or points to evidence, provide reason- (Iwuanyanwu, 2022; Osborne, 2010). Further to that, ing episodes to back up counterclaims, and at best students use the structural elements of arguments such use evidence to support strategy choice, and finally as claims, data, evidence, reason, and counterclaims to formulate reasoned solutions (Erduran & Park, 2023; make, defend, and evaluate judgments. At best, students Iwuanyanwu, 2020). show how one argument’s claim serves as the data for another argument (Iwuanyanwu, 2020) Judge the reasonableness of alternative solutions to Students argue for or against alternative solution Students pose provocative questions about alternatives, resolve uncertainties pathways, critique selected strategies, clear anomalies/ uncertainties, or missing components in order to exam- ambiguities, use evidence/data to cite or insist on their ine alternative solutions within or across small groups positions, particularly after listening to others (Belland and use the sum of all defended reasonable solutions to et al., 2011; Ogunniyi, 2022) reach a consensus (Geifman & Raban, 2015) Defend reasonable solutions/collaborative consensus Students participating at the intra-, inter-, or trans- Students attempt to justify their positions, play devil’s reached locutory argument levels use the argument structure advocate, reconsider their positions in light of more (data, claims, warrant, backing, or other conceptual compelling arguments (Iwuanyanwu, 2017), and, most resources compatible with valid scientific knowledge) importantly, demonstrate how they find solutions to a as well as use text structures such as if/then/but/ given problem (Voss, 2006) therefore to advance reasoned arguments required to produce justified solutions Research in Science Education (2023) 53:1063–1079 1069 a 12-week period. The language of instruction was English. Both groups received the same content (advanced mechanics) and the same amount of teacher contact and were pretested and posttested with the same instrument (4  h in weeks 1 and 12). During the 12-week teaching–learning period, the exp-group participated in a series of argumentation-based lessons delivered by the author using DAI (Erduran & Park, 2023; Gürel & Süzük, 2017; Iwuanyanwu & Ogunniyi, 2020), while the Ctrl-group received expository instruction from another instructor in the science education faculty (Adams & Wieman, 2015). Both teach- ing approaches consisted of cycles of preparation and reflection, which in turn led to the next cycle of instruction (Geifman & Raban, 2015). During week 1 (0.5 h on Day 1) of the study, students were briefed on the study project and given guidelines for their learn- ing behaviors throughout the study. Following this meeting, the two-hour pretest data col- lection session took place. In week 2 (5 min before the DAI lesson began), students were asked to form groups of 5–6 students per group and were given a unique identification code for data collection. Using the DAI guidelines as presented in the literature review, teaching and learning were facilitated through three key phases to actively engage students in learning scientific concepts as individuals (elaborating intra-locutory arguments within individuals), as small groups (elaborating inter-locutory arguments between subgroups), and as a whole class (elaborating trans-locutory arguments across groups). Following this mode, learning phys- ics concepts using basic argument structures was taught to help students learn how to use the structural elements of arguments, such as claims, data, evidence, reason, and counter- claims to formulate reasoned solutions to scientific problems (see Table  1). In weeks 2 through 12, the DAI process was scaffolded, and students were guided. The structure and guidance decrease as students become better at creating evidence-based arguments to solve scientific problems and communicate scientific knowledge. Moreover, since scientific problems by their nature require arguments and counterargu- ments to solve them (Voss, 2006), students in the Ctrl-group received guided instructions from their instructor, who likes to play the role of devil’s advocate. Since both groups received the same content and research instrument, the instructor in the Ctrl-group was asked to guide her students to become aware of the four target variables in order to solve scientific problems and develop reasoned solutions. Her instruction was processed as shown in Table 2. Data Collection and Analysis The instrument (consisting of 17 items) was administered to both the Exp-group and the Ctrl-group as a posttest (within 2 h at week 12) toward the end of the semester. Following posttest data collection, the researcher compiled each student’s solution script, removed identifying information, and then assigned a number code. The same number code was used to track student performance and progress throughout the study. Student performance on the targeted science concepts and variables (Tables 1 and 2, respectively) was then ana- lyzed using one-way MANCOVA (multivariate analysis of covariance), which was consid- ered the most appropriate analysis to examine whether the two study groups differed on the outcomes of the four variables after the intervention. The purpose was to test whether par- ticipation in DAI and expository classroom activities helps students develop valid problem- solving strategies, formulate and defend arguments, and find reasoned solutions to given problems. Using pretest scores as covariates, the outcome variables were assumed to have 1 3 1070 Research in Science Education (2023) 53:1063–1079 Table 2 Expository instruction for scientific problem solving Targeted variables Description of expository activity Construct valid problem-solving strategies Step I: introduces the topic through interactive Formulate defensible arguments for chosen strate- whiteboard (ppt) and communicates the focus of gies the lesson Judge the reasonableness of alternative solutions to Step II: links the new lesson with previous knowledge resolve uncertainties Step III: explains the content of the topic by giving Defend reasonable solutions/collaborative consen- the conceptual and operational definitions of con- sus reached cepts and relevant knowledge resources Step IV: in some cases, instructs the students to check for meaning of concepts Step V: gives and solves a few examples of problems Step VI: calls on a volunteer student who has solved the problem to share the solution with the class To stimulate her students’ reasoning as they think- aloud about scientific phenomena and associated problems to solve, the lecturer was noted prompting her students to consider the following thought- provoking questions: What do I already know about the given problem? How can I construct a problem space that does not only have some levels of plausibility or accept- ability, but is capable of leading to justifiable solution/s? What arguments do I need to produce to convince others why my solution is the best possible solution about the given problem? I know scientific problems may have multiple solutions and often are not final: have I considered other alternatives and examined their constraints? Have I provided evi- dence to back up alternative or proposed solutions? How will I know if the problem has been satisfacto- rily resolved? a linear relationship in terms of multivariate normality and variance–covariance matrices 2 (x = 33.19, df = 28, p = .012) (Creswell, 2013). Findings and Discussion According to the results of this study, under two different teaching approaches, both study groups made significant progress in learning how to create and use evidence-based argu- ments to solve scientific problems and communicate knowledge. For example, the quantita- tive data and student reasoning episodes showed that a significant number of students pro- gressed from the pretest to posttest phase in terms of (a) constructing valid problem-solving strategies, (b) giving defensible arguments about their chosen strategies, (c) judging the reasonableness of alternative solutions to resolve ambiguities, and (d) defending reason- able solutions/collaborative consensus. From pretest to posttest, the Exp-group made sig- nificant progress in constructing valid problem-solving strategies (t = 6.75, p < 0.001) and was able to provide defensible arguments for their choice strategies (t = 4.39, p < 0.001). The Ctrl group made significant progress from pretest to posttest only in developing valid problem-solving strategies (t = 3.26, p < 0.001). 1 3 Research in Science Education (2023) 53:1063–1079 1071 Table 3 Results of one-factor MANCOVA Variables Wilk’s Λ df F p-value Effect size (η2) Construct valid problem-solving strategies Pretest 0.543 3 25.63 0.000*** 0.397 Instructional approach 0.716 3 6.72 0.000*** 0.481 Formulate defensible arguments for chosen strategies Pretest 0.661 3 29.40 0.000*** 0.120 Instructional approach 0.728 3 4.82 0.000*** 0.223 Judge the reasonableness of alternative solutions Pretest 0.46 3 49.56 0.000*** 0.754 Instructional approach 0.656 3 10.67 0.0021* 0.582 Defend reasonable solutions or collaborative consensus Pretest 0.643 3 33.08 0.000*** 0.491(L) Instructional approach 0.765 3 3.40 0.000*** 0.240(M) **p < 0.005 *p < 0.05 ***p < 0.001 Table 4 Results of repeated measure of ANOVA Post hoc Source F p-value Partial eta Time p-value Group square (η2) Time 5.36 0.000*** 0.057 V3 > V1 0.042*** Exp > Ctr (0.042) Group 0.297 0.461 0.043 V4 > V1 0.016*** Time × group 4.82 0.000*** 0.167 V1 construct valid problem-solving strategies, V2 indicate defendable arguments about choice of strategies, V3 judge the adequacy of alternative solutions for uncertainties, V4 defend reasonable solutions or collabo- rative consensus. **p < 0.005 *p < 0.05 ***p < 0.001 Table 3 is a summary of the results for the four variables measured in the two study groups at pre- and posttest levels. A one-way MANCOVA using pretest scores as a covari- ate suggests that instructional approaches (Wilk’s Λ = 0.765, p < 0.001) have a statistically significant impact on postintervention outcomes for the variables tested. The univariate test confirmed significant differences in posttest scores between the study groups on judging the reasonableness of alternative solutions to resolve ambiguities (F = 10.67, p = 0.0021) and defending reasonable solutions/collaborative consensus (F = 3.40, p 0.001). Thus, stu- dents’ posttest scores on defending arguments related to solutions or consensus about sci- entific phenomena were significantly influenced by the instructional approach used. Fur- ther post hoc tests (Tukey HSD), shown in Table 4, indicate that there was a difference between the two study groups at posttest (Exp group > Ctrl group), (p(post-Var.3) = 0.042) and (p(post-Var.4) = 0.016). Note: see Table 4 for a full description of “Var. 3” and “Var. 4.” 1 3 1072 Research in Science Education (2023) 53:1063–1079 The results of repeated measures of ANOVA with one factor Exp or Control are summarized in Table 4. The between-subjects factor is the group, and the within-subjects factor is the time (pre or post scores). Scores on the pretest demonstrated significant differences between time and exp-group (F = 5.36, p 0.001), along with a significant time main effect (F = 4.82, p 0.001). However, posttest scores were not significantly affected by the Ctrl-group (F = 0.290, p = 0.461). A pairwise comparison of test time for learning in the study groups is shown in Table 5. There was a significant difference in posttest scores between study groups (F = 5.36, p = 0.001, η 2p = 0.032). However, the pretest scores were not significantly different between the two groups (F = 4.82, p = 0.195, η 2p = 0.0063). In the Exp-group, the posttest results differed significantly from the pretest by (F = 0.46, p < 0.001, η 2p = 0.02) and in the Ctrl- group by (F = 6.56, p < 0.001, η 2p = 0.26). Also, in the experimental group, the difference between pre- and posttests means was greater than in the control group. Overall student performance suggests that the control group performs at a much lower level than the experimental group on physics problems that depend on theoretical foundations or assumptions. This suggests that encouraging students to learn physics concepts and solve related problems in a DAI-based classroom may help students develop different reasoning skills (Asterhan & Swartz, 2007; Osborne, 2010; Voss, 2006). Of the 46 student teachers in the Exp group, 28 provided evidence-based arguments and rationales for multiple or alternative solutions. By comparison, 11 of their peers in the Ctrl group did likewise. More than 49% of the group could not judge the adequacy of their proposed solutions to ill-structured problems using the structural elements of arguments such as claims/counterclaims, data, evidence, and reasoning. The following is an example of an ill- defined scientific problem that student teachers addressed during the pre–posttests (SP -Q2, Table 6, Figs. 1, 2, 3, 4). Student Teachers’ Deliberation on Scientific Problem When student teachers were presented with the task SP-Q2 during the pre- and posttests, they were asked to perform the first two of the four variables examined in the study. In this phase, they were allowed to make mistakes, correct mistakes, make predictions, and self-regulate their own thinking. In small groups of five, students negotiated the question and its meaning. They gathered data and used available evidence to approach the problem Table 5 Pairwise comparisons of item scores Observed confidence level (95%) Comparisons MD S error p-value Low bond Up bond Posttest Exp–pretest Exp-group 4.45 0.78 0.000*** 2.56 5.95 Posttest Ctrl–pretest Ctrl-group 2.11 0.43 0.000*** 3.58 5.12 Pretest-Exp–pretest-Ctrl − 0.39 2.03 0.195 − 1.83 1.18 Posttest-Exp–posttest-Ctrl 1.95 0.49 0.000*** 5.54 8.09 Exp experimental group, Ctrl control group. **p < 0-0.005 ***p < 0.001 *p < 0.05 1 3 Research in Science Education (2023) 53:1063–1079 1073 Table 6 Inquiry about the Great Pyramid SP-Q2: about 4500 years ago, ancient Egyptians constructed the Great Pyramid, which is made up of about 2,300,000 stone blocks, the majority of which weigh between 2000 and 3000 kg. Many generations have wondered how the ancient engineers and craftsmen were able to carry the stones into place, which are over 140-m high. Some claim that a big group of workers probably lifted the blocks up a substantial earthen ramp that went up one side of the pyramid at a little incline during building. However, there is no proof for this theory (such as debris or painted images). According to modern scholars, the pyramid was encircled by a spiral ramp. Such a ramp, meanwhile, would have been extremely unstable, and moving a 2000-kg stone around it would have been difficult, if not impossible Develop valid problem-solving strategies, formulate defensible arguments for the chosen strategies, and evaluate and defend the reasonableness of alternatives to answer this question: how did the ancient Egyp- tians move the blocks up and into position? Fig. 1 Pyramid scientific ill-defined problem. Source: unsplash.com in their subgroups. After this phase, students were referred to the remaining variables, but this time they were asked to develop possible strategic solutions based on sound arguments and evaluate their solutions. To do this, they planned, collected, recorded, and analyzed the approach they would take to address the problem. To complete the final phase, each individual or group was asked to explain their proposed solutions, applying the argumenta- tion skills they had acquired. As each group shared their progress, they were encouraged to challenge the evolving evidence by making claims/counterclaims, presenting their evi- dence, and communicating and justifying their solutions. Using the guidelines outlined in 1 3 1074 Research in Science Education (2023) 53:1063–1079 Fig. 2 Initial predictions of work done during the construction of the pyramid Fig. 3 Final predictions of work done during the construction of the pyramid Fig. 4 Free-body diagram to sup- port arguments generated 1 3 Research in Science Education (2023) 53:1063–1079 1075 Table  1, the instructor constantly challenged the Exp-group to provide reasons for their claims/counterclaims, to pose thought-provoking questions, to persuade their classmates, who serve as critical friends and to provide evidence or data to support or refute opposing viewpoints. Comparison of Student Teachers’ Responses to Ill‑Defined Problems SP‑Q2 Five student teachers in the Ctrl-group reported closely related problem-solving strategies and some arguments that included text structures such as if/then/but/therefore to support their solutions to Task SP-Q2 (Table 6, Fig. 1), as did most of their peers in the Exp-group. However, the extent to which they presented arguments in support of their solutions based on the four outcome variables differed markedly from their peers in the DAI-based class, who consistently presented better arguments and used reasoned evidence to support their solutions. Considering that a problem solver may frame ill-structured problem-solving strategies differently based on his or her knowledge, experience, and/or insight into the problem context, the final solutions produced by the Ctrl group showed some inconclusive reasoning episodes. According to the final proposed solutions of SP-Q2 presented by the Ctrl-group, some incompatible excerpts of invalid scientific knowledge found in their solu- tion scripts suggest they considered the problem statement superficially without adequately defining the problem context or exploring it using reasoned arguments. Their lack of expe- rience with argumentation instruction may account for the discrepancy. By engaging stu- dents in argumentation instruction while learning physics concepts, they can develop com- plex learning skills that will assist them in their future workplaces and in everyday life (Iwuanyanwu, 2022). Due to space limitations, only a few subgroup presentations were selected to show how students developed problem-solving strategies, made and defended arguments, and developed reasoned solutions to item SP-Q2. The question was as follows: how did the ancient Egyptians move the blocks up and into position when building the Great Pyramid? For simplicity, a student teacher is referred to as ‘ST.’ Excerpt 1: Subgroup 3 Presented, Subgroup1 Responded ST33: The question says: …in building the Great Pyramid, how did the ancient Egyptians move the blocks up and into position? ST77: The Pyramid has a shape of a prism block…how they moved the blocks up is a mystery… ST33: …but we can’t simply say the problem belongs to mystery… ST49: I have been looking at the Pyramid image, I think they probably moved the blocks up the side of the Pyramid using a rope… ST61: How is that possible? …how did they build the side of the Pyramid then, to now use it to support the upward movements of the blocks? ST49: …they probably started from the foundation to get to a height that requires to pull the blocks up and into position ST49: Here is my sketch, the rope is parallel by the opposite link… ST33: …they probably had no machines, like crane, so those men pulling the rope must be able men to pull 2000–3000kg block up there against gravity (Fg) and fric- tional force (ff )…how many men could have done so? ST61: …certainly, it will depend on how strong the men were…and the magnitude of force each man could produce 1 3 1076 Research in Science Education (2023) 53:1063–1079 ST77: That I agree, even so one wonders how they managed to secure the block against gravity and frictional force that Mercy was alluding to… ST49: … with the frictional force downward the plane to oppose the pending motion… one can say from Newton’s second law that F − mgsin − fs ST61: The block could have been secured to a wood sled or something and …is pulled by multiple ropes ST49: So, are you saying that one rope…a strong rope will not do the job? ST77, ST33: …definitely not, Dawson …think about 2000-3000kg… ST49: Therefore, with multiple ropes once the block began to move upward, one expects.F = smgcos + mgsin ST77: When the block is on the verge of moving the pyramid side, the static friction: fs = fs max ST61: It makes sense then from this ST33: So, the blocks could have been pulled up into position by teams of able men as Dawson said from the beginning…no evidence is available to suggest exactly how many men did so ST49: Yeah For student ST61, the premise presented by Dawson (ST49) suggesting that “the build- ers of the Great Pyramid probably used ropes to move the blocks up the side of the pyra- mid” (ST49) led him to ask, “…how did they build the side of the pyramid then to use it now to support the upward movements of the blocks?” His search for warrant was sim- ply to point out that ‘facts’ are the arguments that can be made to support the premise. However, to make sure he understood the context, ST49 tacitly constructed his initial prob- lem strategies and asked his group to identify and explore possible limitations. Available research suggests that ST49 is aware of the activity at levels of different cognitive styles, prior knowledge, experience, and reasoning ability (Jonassen, 2011; Osborne, 2010), and it is this level of awareness of the activity that is most likely to have an impact on him (Redish & Kuo, 2015) and lead to internalization (Adams & Wieman, 2015). As can be seen, their arguments relate to important components of the kinematics and dynamics of the movements of objects. This supports the assumption that it would have required teams of capable men to lift 2000 to 3000 kg blocks of stone against gravity and the force of friction and move them into position, since machines such as cranes did not exist in those years (descriptive claim of ST33). Since there is much to consider, and after refuting her colleague’s initial claim that “getting the blocks up and into position is a mys- tery” (ST77), she colored her thoughts with Newtonian concepts—how many men did it take? In response, ST61 added that it certainly depends on the strength and power that each man can muster. In fact, as part of the solution path, his argument holds that “… several ropes could have been used to move the blocks upward, which explains the connection between epistemological optimism and the later inclusion of “ F = Smgcos + mgsin .” After looking at some major components of the problem, they concluded that “…the blocks could have been pulled into position by teams of capable men …but in terms of the hard core, there is no indication of exactly how many men did this.” According to the findings of this study, approximately 27 students who received DAI lessons and 11 students who received expository lessons completed 9 of the 17 tasks at a level that met all four dependent variables tested. The remainder from both groups were unable to dismiss the potential conflicts inherent in the ill-defined structured tasks 1 3 Research in Science Education (2023) 53:1063–1079 1077 as unimportant or the whim of mythical science. Although Shin et al. (2003) argued that students can solve unstructured science problems using general strategies if they are aware of or have sufficient knowledge about the problem domain, the current study adds that in order to solve unstructured problems and formulate reasoned solutions, students must be able to formulate and defend arguments that support their solution paths/strategies. In terms of initiating and reflecting on alternatives, constructing arguments, evaluating, and reasoning, most students were idiosyncratic on some of the unstructured problems in the area of force and motion as they occur in everyday life. They were unable to identify with the meanings articulated by their classmates, who serve as critical friends. Other stud- ies have shown that successful solvers of unstructured problems need to justify their deci- sion about selected strategies in order to generate plausible or acceptable solutions (Bel- land et al., 2011; Gürel & Süzük, 2017; Syafril et al., 2021). While some of the students in the current study attempted to justify the decisions that led to the formulation of problem- solving strategies, some others did not do so satisfactorily. As a result, this study shows that the creation and defence of arguments to develop a reasoned solution to an unstructured scientific problem occurs through processes controlled by conditions that include students’ existing knowledge, reflective judgement, and subject-specific knowledge. Although not all student teachers were able to integrate expertise arguments and reasoning skills when they were needed to construct valid problem-solving strategies, the arguments and justifications generated from the completed worksheets indicated that they were motivated to engage in the activity as critical friends (e.g., ST11, ST27, and ST8). Conclusion This study has described an interesting type of science learning that fuses explicit teach- ing of science knowledge with student solving of unstructured problems. It has also pro- vided compelling data showing how dialogic argumentation instruction can help foster stu- dents’ thinking when they must make and defend judgments about different sides of issues, thereby contributing to the extant literature (Erduran & Park, 2023; Evagorou & Osborne, 2013; Tarekegn et  al., 2022). The results of the current study are consistent with other research evidence demonstrating the efficacy of dialogic argumentation instruction for cul- tivating scientific knowledge among students and teachers (Asterhan & Schwarz, 2007; Iwuanyanwu & Ogunniyi, 2020; Lubben et al., 2010; Ogunniyi, 2022). Additionally, the current study supports the claims of Erduran and Park (2023) and Gürel and Süzük (2017) that argumentation instruction plays a more prominent role in assisting student teachers to better understand physics concepts and solve related prob- lems. To some extent, the use of dialogic argumentation instruction appears to have helped students in the Exp-group mobilize better strategies for solving scientific problems. For example, the underlying assumptions they made about the four variables listed in Table 1 all showed indicators of shared tacit knowledge, sharing of cognitive discourse and under- standing among their classmates. In addition, the group was generally enthusiastic about the intervention program delivered via DAI, except for the few who found the repetition of the stages of the four variables tedious when solving problems individually. Nonetheless, students need argumentation skills and knowledge to solve complex problems such as those encountered in everyday life. Finally, this study has shown that engaging student teachers in dialogic argumentation instruction can help them improve their ability to think, reason, and solve different types 1 3 1078 Research in Science Education (2023) 53:1063–1079 of problems related to subject knowledge and socio-scientific contexts, which are essential skills for the science teaching profession. It should be noted that the results of this study may be affected by some limitations. Primarily, this was a case study of physics student teachers from a single institution taught by different instructors. Generalizations of the results to the entire population of university science students should be made with caution. For example, the study lasted only 12 weeks and included only 79 students. Therefore, a study with larger samples is recommended before implementing the approach on a large scale. Funding Open access funding provided by North-West University. Declarations Conflict of Interest The author declares no competing interests. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com- mons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creati vecom mons.o rg/ licen ses/b y/4. 0/. References Adams, W. K., & Wieman, C. E. (2015). Analyzing the many skills involved in solving complex physics problems. American Journal of Physics, 83(5), 459–467. https://d oi. org/ 10.1 119/1. 491392 3 Asterhan, C., & Schwarz, B. (2007). The effect of monological and dialogical argumentation on concept learning in evolution theory. Journal of Education Psychology, 99(3), 626–639. Belland, B. R., Glazewski, K. D., & Richardson, J. C. (2011). Problem-based learning and argumentation: Testing a scaffolding framework to support school students’ creation of evidence-based arguments. Instructional Science, 39(5), 667–694. Creswell, J. W. (2013). Qualitative inquiry and research design: Choosing among five approaches (3rd ed.). Sage. Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual Review of Psychology, 53(1), 109–132. Erduran, S., & Park, W. (2023). Argumentation in physics education research: Recent trends and key themes. In M. F. Taşar & P. R. L. Heron (Eds.), The international handbook of physics education research: Learning physics (pp. 16–32). AIP Publishing. Erduran, S., Kaya, E., & Çetin, P. S. (2016). Pre- service teachers’ perceptions of argumentation: Impact of a teacher education project in Rwanda. Boğaziçi University Journal of Education, 33(1), 1–21. Etkina, E., Brookes, D., & Planinsic, G. (2019). Examples of ISLE-based learning of traditional physics topics and examples of ISLE-based physics problems. In E. Etkina, D. Brookes, & G. Planinsic (Eds.), Investigative science learning environment: When learning physics mirrors doing physics (pp. 2.1– 2.32). IOP Publishing. Evagorou, M., & Osborne, J. (2013). Exploring young students’ collaborative argumentation within a socio- scientific issue. Journal of Research in Science Teaching, 50(2), 209–237. Geifman, D., & Raban, D. R. (2015). Collective problem-solving: The role of self-efficacy, skill, and prior knowledge. Interdisciplinary Journal of e-Skills and Lifelong Learning, 11, 159–178. https:// doi.o rg/ 10.2 8945/2 319 Ghebru, S., & Ogunniyi, M. (2017). Pre-service science teachers’ understanding of argumentation. African Journal of Research in Mathematics, Science and Technology Education, 21(1), 49–60. Gürel, C., & Süzük, E. (2017). Pre-service physics teachers’ argumentation in a model rocketry physics experi- ence. Educational Sciences: Theory & Practice, 17, 83–104. https:// doi. org/ 10.1 2738/e stp. 2017.1.0 042 Hansson, L., & Leden, L. (2016). Working with the nature of science in physics class: Turning ‘ordinary’ classroom situations into nature of science learning situations. Physics Education, 51(5), 1–6. 1 3 Research in Science Education (2023) 53:1063–1079 1079 International Council for Science. (2011). Report of the ICSU ad-hoc review panel on science. Paris, France: Available from http://w ww.i csu. org/p ubli cation s/ repor ts- and- review s/e xter nal-r eview-o f- icsu. Iwuanyanwu, P. N. (2017). An analysis of pre-service teachers’ ability to use a dialogical argumentation instructional model to solve mathematical problems in physics. Unpublished Doctoral Thesis. Cape Town: University of the Western Cape. Iwuanyanwu, P. N. (2019). Students understanding of calculus-based kinematics and the arguments they generated for problem solving: The case of understanding physics. Journal of Education in Science, Environment and Health, 5(2), 283–295. https:// doi.o rg/1 0.2 1891/j eseh. 581588 Iwuanyanwu, P. N. (2020). Nature of problem-solving skills for 21st century STEM learners: What teachers need to know. Journal of STEM Teacher Education, 55(1), 27–40. https:// doi.o rg/1 0. 30707/ JSTE55.1 / MMDZ8 325 Iwuanyanwu, P. N. (2022). What Students Gain by Learning through Argumentation. International Journal of Teaching and Learning in Higher Education, 34(1), 97–107. Iwuanyanwu, P.N. & Ogunniyi, M.B. (2018). Scientific and indigenous worldviews of pre-service teachers in an interactive learning environment. In Proceedings of the 4th Annual Conference of the African Asso- ciation for the Study of Indigenous Knowledge Systems. Moshi, Tanzania: Ngā Pae o te Māramatanga. Iwuanyanwu, P. N., & Ogunniyi, M. B. (2020). Effects of dialogical argumentation instructional model on pre-service teachers’ ability to solve conceptual mathematical problems in physics. African Journal of Research in Mathematics, Science and Technology Education, 24(1), 121–141. https:// doi. org/1 0.1 080/ 18117 295.2 020. 17483 25 Jonassen, D. H. (2011). Learning to solve problems: A handbook for designing problem solving learning environments. Routledge. Koichu, B., Schwarz, B.B., Heyd-Metzuyanim, E., Tabach, M., & Yarden, A.(2022). Design practices and principles for promoting dialogic argumentation via interdisciplinarity. Learning, Culture and Social Interaction. 37:100657. https:// doi.o rg/ 10. 1016/j.l csi.2 022. 100657. Kuhn, D., & Udell, W. (2007). Coordinating own and other perspectives in argument. Thinking and Reason- ing, 13(2), 90–104. Lubben, F., Sadeck, M., Scholtz, Z., & Braund, M. (2010). Gauging students’ untutored ability in argumen- tation about experimental data: A South African case study. International Journal of Science Educa- tion, 32(16), 2143–2166. https:// doi.o rg/ 10. 1080/ 095006 9090 33318 86 Ogunniyi, M.B. (2022). Implementing a socioculturally relevant science curriculum: The South African experience. In: Atwater, M.M. (eds) International handbook of research on multicultural science edu- cation (pp. 819 -837). Springer International Handbooks of Education. Springer, Cham. https://d oi.o rg/ 10.1 007/ 978-3- 030-8 3122-6_ 31. Osborne, J. F. (2010). Arguing to learn in science: The role of collaborative, critical discourse. Science, 328, 463–466. Osborne, J. F. (2019). Not “hands on” but “minds on”: A response to Furtak and Penuel. Science Education, 103, 1280–1283. https://d oi. org/1 0. 1002/s ce. 21543 Redish, E. F., & Kuo, E. (2015). Language of physics, language of math. Science and Education, 25(5–6), 561–590. Schreiber, L. M., & Valle, B. E. (2013). Social constructivist teaching strategies in the small group class- room. Small Group Research, 44(4), 395–411. https://d oi.o rg/ 10. 1177/ 10464 964134 88422 Shin, N., Jonassen, D. H., & McGee, S. (2003). Predictors of well-structured and ill-structured problem solving in an astronomy simulation. Journal of Research in Science Teaching, 40(1), 6–33. Syafril, S., Latifah, S., Engkizar, E., Damri, D., Asril, Z., & Yaumas, N. E. (2021). Hybrid learning on prob- lem-solving abilities in physics learning: A literature review. Journal of Physics: Conference Series, 1796(1), 012021. https:// doi. org/1 0.1 088/1 742- 6596/ 1796/1/ 012021 Tarekegn, G., Osborne, J., & Tadesse, M. (2022). Impact of dialogic argumentation pedagogy on grade 8 students’ epistemic knowledge of science. In M. Kalogiannakis, & M. Ampartzaki (Eds.), Advances in research in STEM education. IntechOpen. https://d oi. org/ 10. 5772/i ntec hopen. 104536. Toma, R. B., & Lederman, N. G. (2020). A comprehensive review of instruments measuring attitudes toward science. Research in Science Education, 1–16. https://d oi. org/ 10. 1007/s 11165- 020-0 9967-1. Toulmin, S. (2003). The uses of argument. Cambridge University Press. UNESCO. (2020). Policy brief. Education during Covid-19 and beyond http://u nesd oc. unesco.o rg Voss, J. F. (2006). Toulmin’s model and the solving of ill-structured problems. In D. Hitchcock & B. Verheij (Eds.), Arguing on the Toulmin model: New essays in argument analysis and evaluation (pp. 303–311). Springer. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 1 3