Merging Oral and Written Argumentation: Supporting Student Writing Through Debate and SRSD in Inclusive Classrooms

Abstract

Argumentation is a key competence (as emphasized by the European Union) for lifelong learning and democratic participation. Written argumentation is a challenging task for students, and to date, no study has investigated the impact of a combined teaching approach of debating and SRSD writing lessons on written argumentation skills. This study addresses this gap by linking debating and SRSD writing lessons for the first time in grade 9 classrooms, employing a debating format and the Self-Regulated Strategy Development (SRSD) approach. We aimed to assess the impact of the combination of debating and SRSD writing lessons on the quality of students’ argumentative writing, examining text quality across school types (higher and lower academic track) and intervention sequences (debating–SRSD and SRSD–debating). This quasi-experimental study included 357 ninth-grade students from six rural and urban German schools, split between higher (57.9%) and lower (42.1%) academic tracks. Over four measurement points, the students participated in randomized debating or SRSD-based writing lessons, each comprising six 90-minute sessions conducted by trained coaches from the research team. Text quality was measured through standardized writing tasks and the double-blinded rating of text and language pragmatics quality on a six-point scale. Both interventions improved text quality, with significant gains observed in the post-tests, and gains were especially visible for lower-track students. This study demonstrates the effectiveness of linking debating and SRSD writing lessons in improving writing skills. Our findings support the inclusion of debating as a motivational precursor to writing, emphasizing adaptability in teaching strategies to accommodate diverse student needs. We recommend confirming these findings and informing broader curricular reforms in further research.

1. Introduction

Inclusion entails all individuals having the opportunity to develop their potential and participate fully in society (European Agency for Special Needs and Inclusive Education, 2025). In the educational context, this means engaging all students with shared topics in the classroom through individualized and differentiated approaches (UNESCO, 2020). The overarching goal is to ensure that every student advances to the next stage of their educational journey and develops literacy skills—without excluding anyone (European Agency for Special Needs and Inclusive Education, 2025). Developing literacy skills, including argumentative abilities, helps prepare students for independent living. These skills are useful in situations such as negotiating contracts, seeking credit, or addressing workplace conflicts.

Formulating convincing arguments is also one of the “Attitudes and Values” focused on by the OECD for developing communication skills while addressing global challenges (OECD, 2019, p. 5) and belongs to the “UNESCO Global Education Agenda 2030” (UNESCO, 2023, p. 28).

In general, argumentative writing places intense demands on secondary-school students, who, in the majority of cases, are either required to imagine their dialogue partners within the written argumentation or are given a fictitious argumentation context. This is one of the primary reasons they often find argumentative writing more difficult than oral argumentation. In addition, it is a particularly demanding metacognitive activity in which the writers must invest cognitive resources in representing their arguments and those of their imagined dialogue partners (Ferretti & Graham, 2019; Hayes, 2012). The challenge for students arises from the fact that they must also consider the arguments for the opposing position when composing written material (Felton et al., 2015; Wolfe et al., 2009; Leitão, 2003).

In addition to generating arguments and the corresponding support, the main difficulty lies in relating these arguments to each other and using them for one’s own argumentation. Furthermore, argumentative writing involves formulating claims; supporting them with examples and evidence; and combining claims, reasoning, and situational context to develop arguments (Mombaers et al., 2024). The Self-Regulated Strategy Development (SRSD) approach provides a structured framework of writing strategies designed to enhance argumentative skills, particularly among students who experience difficulties with writing.

SRSD constitutes one of the most-studied approaches to the design of writing lessons in different text genres (Graham et al., 2023, 2022; Sun et al., 2022; Graham & Perin, 2007), and it is a strategy-oriented, multimodal, evidence-based approach to teaching writing and self-regulation skills. With SRSD, writers should be able to master and independently regulate the (meta-)cognitive processes related to writing. Self-regulated writing refers to self-initiated thoughts, feelings, and actions that individuals use to achieve literacy goals (Hayes, 2012; Zimmerman, 2002).

First, this includes criterion-based progression in six phases (not six lessons), in which the students receive a great deal of support in developing their skills at the beginning, with the level of support decreasing as the series of lessons progresses (see Table 1).

Second, instructions on how to write content such as a pro-and-con argument are frequently repeated during the series of lessons as part of the memorization process to consolidate the students’ knowledge and predictions. This is carried out, for example, through mnemonic strategies, using verbal and visual mnemonics (Malpique & Simăo, 2019).

Another key principle of SRSD is peer knowledge and feedback: students work in teams to write essays and exchange them for feedback based on set criteria.

SRSD has also been proven to be effective with respect to writing argumentative material. De La Paz (2005) explored the effects of combining historical knowledge and argumentative writing skills through SRSD with eighth-grade students (n = 70), having them write argumentative texts on historical issues. The results showed that the intervention group produced more accurate, persuasive, and longer texts with more arguments than the control group (n = 62). Moreover, Mason et al. (2013) found that students (7th and 8th grade) who received SRSD instruction alongside regular lessons produced more convincing and longer argumentative texts than those in comparison groups. A similar study involving 5th and 6th graders (n = 564) also showed that students in an SRSD intervention group wrote better and longer argumentative texts than those in the control group (n = 266) (Mason et al., 2017).

In addition to developing writing skills, strong argumentation abilities, such as debating, are also essential. A multimodal approach to argumentation is required in order to support learners both in writing and orally. Combining both modes enables deeper learning and understanding of how arguments can be used and how they work in oral debates, as the debate is simulated but represents a real dialogue. Written argumentation, on the other hand, also requires taking the perspectives of both the pro and con sides, but is delayed by the longer planning, writing and revising of thoughts and does not have a direct addressee. Rather, the metacognitive demands of writing are very high and challenging for students without sufficient prior knowledge.

In order to appeal to the different strengths and senses of the students, the intervention addresses different modes of argumentation. In both debating and written argumentation, students must research information and prepare it for the debate or text. While students convey arguments in verbal interaction through the use of spoken language, facial expressions and gestures in a debate, arguments in pro and con essays are formulated in writing, whereby texts can also form the basis for verbal interactions. The same applies in reverse to debates, which can be written down.

While SRSD teaches students explicit strategies for planning, organizing and revising argumentative texts, it does not itself offer interactive, dialogical practice in testing and defending claims against real opponents. Debating, on the other hand, promotes precisely this skill by placing argumentation in a dynamic, social context in which students must anticipate counterarguments, respond spontaneously and refine their reasoning. Debating and SRSD are therefore complementary: SRSD provides a structured framework for self-regulated writing, while debating provides the authentic argumentation practice that motivates and contextualizes these strategies. By integrating both approaches, students can transfer their skills between oral and written expression, thereby strengthening both the cognitive and social dimensions of argumentation.

Debating is an effective method for building argumentation skills, facilitating linguistic, social, and democratic learning as participants exchange diverse perspectives in a structured format (Achour et al., 2020; Jagger, 2013). In contrast to other forms of dialogue, debates follow a predetermined structure in the form of opening, free, and closing speeches, the roles of which are underlined by teaching debating in this study. The pioneering framework for school debates in Germany is the Jugend debattiert (Youth Debates, transl. by the authors) format (Kemmann et al., 2022; Hielscher et al., 2019). Debates in this format are conducted with two students on the pro side and two on the con side, the positions of which are decided before the discussion. In addition, the purpose of a debate is not to reach a consensus but to differentiate one’s own position from that of the opposing side in a well-founded manner.

Early experimental studies on debates and their influence on argumentative writing, like the one conducted by Green and Klug (1990), revealed that debating positively influences the quality of students’ essays. They discovered a significant improvement in the quality of the ‘before’ and ‘after’ essays written by students who participated in debates. At the primary-school level, Malloy et al. (2020) studied debating in social studies using three groups of fifth-grade students (n = 40), who participated in three debate cycles (with 2 or 3 weeks per cycle). Analysis of these students’ writing samples using a four-point scale showed improvement in text quality (e.g., focus, organization, evidence, and language) over the three debate cycles, revealing the positive effect of debating on written argumentation (Malloy et al., 2020).

At the secondary-school level, Majidi et al. (2021) confirmed the validity of this positive relationship between debating and written argumentation. They investigated the effect of debate training on written argumentation in L2 education at three secondary schools (n = 147). In this context, the participants in the intervention groups took part in ten debates without receiving instructions on the general structural aspects of argumentation. Nevertheless, the students in the intervention groups exhibited greater structural and reasoning quality in their arguments than the control group (Majidi et al., 2021). The studies mentioned above indicate that debating has a positive effect on oral and written argumentation.

Firetto et al. (2018) confirmed similar findings, showing that small-group discussions positively influenced the quality of fifth-grade students’ written argumentation. Casado-Ledesma et al. (2021) found that deliberate discussions were suitable for improving the quality of integration regarding argumentative material written by Spanish secondary-school students. These findings confirm the results of the multi-year study conducted by Crowell and Kuhn (2014) in middle schools; they examined the influence of dialogue-based argumentation on the development of cognitive argumentative skills, such as the ability to consider counterarguments (Crowell & Kuhn, 2014; Kuhn & Crowell, 2011). Their findings show that the dialogue-based approach investigated also seems suitable for improving argumentative essays and argumentative knowledge (Lin et al., 2020).

While prior research identifies critical gaps in the academic writing of secondary students (e.g., inconsistent strategy use, low transfer of skills to argumentative tasks), these gaps are particularly pronounced for students from marginalized backgrounds—creating an inequality in access to high-quality writing instruction that limits their academic and postsecondary opportunities. In Germany, structural inequalities due to different school curricula between school types become particularly apparent in Year 9, when pupils on the higher academic track have already spent a year practicing writing pros and cons, while pupils on the lower academic track are only just becoming familiar with this. This discrepancy means that students from lower-track schools often face greater challenges in developing argumentative writing skills, which are required for their final examinations in Year 10. Eliminating this imbalance is not only a matter of educational equity, but also essential for preparing all pupils for democratic participation. Linking the Self-Regulated Strategy Development (SRSD) model with debate instruction offers a promising solution: SRSD explicitly teaches the planning, regulation, and revision strategies students need for written argumentation, while debates provide a motivating and dialogic context in which these strategies can be practiced and transferred.

The aim of this study is therefore to investigate how combining SRSD-based writing instruction with structured debate formats can improve the quality of argumentative writing among ninth-grade students across school types and intervention sequences, thereby addressing both the documented difficulties of adolescent writers and the inequality of access to high-quality argumentation instruction in German secondary schools.

2. Materials and Methods

2.1. Research Questions and Hypothesis

This current quasi-experimental intervention study is the one of the first to focus on promoting written argumentation skills by linking debating and written argumentation training via the SRSD approach. The central aim is to investigate how the text quality of argumentative essays can be improved by implementing two linked debating and SRSD lesson series in grade 9 at six higher- and lower-academic-track schools in Germany. Thus, this study also focuses on students with different starting points and the order of debating and SRSD intervention. Answering the first research question is relevant because the pilot study first collects baseline results that provide an overview of students’ levels of competence in written argumentation. Within the framework of the pilot study, this can be examined for students with different starting points based on their type of school. The following research questions (RQs) indicate the primary focuses of this study (Giera, 2025):

RQ₁: 

Under debating and SRSD intervention conditions, how will the quality of the participants’ argumentative writing develop throughout the study, based on whether they are in the intervention group or the control group?

RQ₂: 

How does school type (lower and higher academic track) affect students’ development of argumentative writing quality under the influence of debating and SRSD intervention conditions?

RQ₃: 

Which order (debating–SRSD or SRSD–debating) is more effective for developing argumentative writing skills?

The results of the analysis of previous studies on oral and written argumentation given in Section 1 can be summarized as follows: First, debating enhances critical thinking, the ability to weigh arguments, and the ability to simultaneously perform linguistic and democratic activities because of its structured format and the nature of confrontation with opposing arguments (Majidi et al., 2021; Malloy et al., 2020; Achour et al., 2020; Hirvela, 2017). This framework of debating may also have an impact on writing argumentative texts. Second, studies conducted on individuals at various levels show that interventions based on SRSD can potentially improve the quality of argumentative texts (Mason et al., 2017, 2013; De La Paz, 2005). Based on these empirical findings (see Section 2), we formed the following hypotheses (H₁–H₃) regarding the research questions above (in brackets):

H₁: 

First, we assume that written argumentation training based on SRSD (Mason et al., 2017; De La Paz, 2005) and debate training (Majidi et al., 2021; Malloy et al., 2020) will help 9th-grade students write superior comprehensive pro-and-con-arguments (RQ₁). Students included in both intervention groups will outperform the control group in terms of text quality.

H₂: 

Second, we assume that the students attending schools with a higher academic track (HSS) will write better essays than those from schools with a lower academic track (LSS). Due to the curricula, HSS students start writing pro-and-con-arguments one year earlier than LSS students. Therefore, the LSS students have less knowledge and experience than HSS students regarding this type of writing. Nevertheless, based on the results of previous SRSD studies (Graham & Harris, 2017; Graham & Perin, 2007), we predict that the students attending lower-academic-track schools will benefit more from the intervention than the HSS students (RQ₂).

H₃: 

Third, we assume that the order of interventions (debating–SRSD or SRSD–debating) will play a role in the student’s ability to develop writing skills. This hypothesis agrees with previous findings (Majidi et al., 2021; Malloy et al., 2020) suggesting that oral argumentation should be taught before writing for didactical and motivational reasons (RQ₃).

2.2. Research Design and Participants

To investigate the research questions and prove these hypotheses, a quasi-experimental design featuring switching replications (Rogiers et al., 2021; Bouwer et al., 2018; Shadish et al., 2002) for the intervention groups and measurements conducted on four occasions was implemented (see Table 1).

Table 1. The course of the project for all the groups.

	Panel WD	Panel DW
Phases	First Intervention W: Written Argumentation (SRSD Approach; see Section 1)	First Intervention D: Debating (Jugend dabattiert; see Section 1)	Control Group: Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Pre-test (t1)	Writing Task A (Appendix A.1), Duration: 60 min	Writing Task A (Appendix A.1), Duration: 60 min	Writing Task A (Appendix A.1), Duration: 60 min
Intervention step 1
Lesson 1	“Develop Background Knowledge”	“Trying to debate” Introduce the rules and structure of debating (e.g., two pro and two con debaters)	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 2	“Discuss it”	“Getting to the point” Prepare an opening speech	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 3	“Model it”	“Referring to each other” Prepare for free discussion	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 4	“Memorize it”	“Drawing conclusions” Prepare a closing speech	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 5	“Support it”	“Clarifying preconditions” Conduct research in preparation for the debate	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 6	Independent Performance	Training debates	Regular lessons
Intermediate Test (t2)	Writing Task B (Appendix A.2), Duration: 60 min	Writing Task B (Appendix A.2), Duration: 60 min	Writing Task B (Appendix A.2), Duration: 60 min
Intervention step 2, groups WD and DW swap interventions.
Step 2 Intervention	Intervention D (like step 1)	Intervention W (like step 1)	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Post-test (t3)	Writing Task C (Appendix A.3), Duration: 60 min	Writing Task C (Appendix A.3), Duration: 60 min	Writing Task C (Appendix A.3), Duration: 60 min
Maintenance Test (t4)	Writing Task D (Appendix A.4), Duration: 60 min (8 weeks post intervention)	Writing Task D (Appendix A.4), Duration: 60 min (8 weeks post intervention)	Writing Task D (Appendix A.4), Duration: 60 min

Table 1. The course of the project for all the groups.

	Panel WD	Panel DW
Phases	First Intervention W: Written Argumentation (SRSD Approach; see Section 1)	First Intervention D: Debating (Jugend dabattiert; see Section 1)	Control Group: Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Pre-test (t1)	Writing Task A (Appendix A.1), Duration: 60 min	Writing Task A (Appendix A.1), Duration: 60 min	Writing Task A (Appendix A.1), Duration: 60 min
Intervention step 1
Lesson 1	“Develop Background Knowledge”	“Trying to debate” Introduce the rules and structure of debating (e.g., two pro and two con debaters)	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 2	“Discuss it”	“Getting to the point” Prepare an opening speech	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 3	“Model it”	“Referring to each other” Prepare for free discussion	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 4	“Memorize it”	“Drawing conclusions” Prepare a closing speech	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 5	“Support it”	“Clarifying preconditions” Conduct research in preparation for the debate	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Lesson 6	Independent Performance	Training debates	Regular lessons
Intermediate Test (t2)	Writing Task B (Appendix A.2), Duration: 60 min	Writing Task B (Appendix A.2), Duration: 60 min	Writing Task B (Appendix A.2), Duration: 60 min
Intervention step 2, groups WD and DW swap interventions.
Step 2 Intervention	Intervention D (like step 1)	Intervention W (like step 1)	Regular lessons, with teachers asking students to practice argumentative writing with examination tasks and no principles of debating or SRSD approach
Post-test (t3)	Writing Task C (Appendix A.3), Duration: 60 min	Writing Task C (Appendix A.3), Duration: 60 min	Writing Task C (Appendix A.3), Duration: 60 min
Maintenance Test (t4)	Writing Task D (Appendix A.4), Duration: 60 min (8 weeks post intervention)	Writing Task D (Appendix A.4), Duration: 60 min (8 weeks post intervention)	Writing Task D (Appendix A.4), Duration: 60 min

The intervention groups were subjected to two lessons series: one designed to improve oral argumentation skills (debating, intervention D), and another based on the SRSD approach, focusing on written argumentation skills (written pro-and-con-argumentation, intervention W). Both interventions included six 90-minute lessons. After these six lessons, the experimental groups swapped interventions. This resulted in two intervention groups, DW and WD, and a control group that was not subjected to either intervention and participated only in regular lessons. Table 1 shows the timeline of this study and the thematic focuses of the two-lesson series for panels DW and WD, along with the activities carried out by the control group. This table also describes when the measurements were conducted and writing tasks (the entire tasks are shown in Appendix A.1, Appendix A.2, Appendix A.3 and Appendix A.4). Table 1 provides an overview of the course of the project for all groups, especially with respect to the structure and topics of the two intervention conditions:

A quasi-experimental design was used because existing classes could not be randomly assigned. In consultation with cooperating teachers, classes were allocated to intervention or control groups, while ensuring balanced ratios. This study took place with ninth graders from six schools in Brandenburg, representing both urban and rural areas. In Germany, after elementary school, students follow different educational pathways: Higher Secondary Schools (HSSs) lead to a university entrance qualification in grade 12, while Lower Secondary Schools (LSSs) end with an intermediate certificate in grade 10. Three HSSs and three LSSs agreed to participate, resulting in a sample of 357 students across 16 classes, of which 10 were intervention groups and 6 were controls. Altogether, 614 essays from HSS students and 447 from LSS students were collected (see

Table 2. Means and standard deviations of MOI scores according to group, school type, and time point.

	t1	t2	t3	t4
Group (all)	M (SD) n = 311	M (SD), n = 304	M (SD), n = 298	M (SD), n = 297
Debate–Writing (LSS)	2.45 (0.58) n = 52	3.12 (0.65) n = 50	3.30 (0.60) n = 49	3.45 (0.62) n = 49
Writing–Debate (LSS)	2.40 (0.59) n = 52	2.98 (0.60) n = 51	3.25 (0.61) n = 50	3.40 (0.63) n = 50
Control (LSS)	2.50 (0.57) n = 50	2.55 (0.56) n = 49	2.65 (0.58) n = 48	2.70 (0.59) n = 48
Debate–Writing (HSS)	2.65 (0.54) n = 54	3.20 (0.60) n = 53	3.50 (0.62) n = 52	3.65 (0.64) n = 52
Writing–Debate (HSS)	2.60 (0.55) n = 53	3.10 (0.58) n = 52	3.45 (0.61) n = 51	3.60 (0.63) n = 51
Control (HSS)	2.70 (0.52) n = 50	2.75 (0.51) n = 49	2.80 (0.53) n = 48	2.85 (0.55) n = 47

Note. M = Mean; SD = Standard Deviation. MOI = Mean Overall Impression. Each score represents aggregated rater judgments. LSS = Lower Secondary School; HSS = Higher Secondary School. n = Number of students contributing data per group.

).

Within intervention classes, students were randomly divided into two subgroups following different intervention sequences (see Table 1). Each subgroup was taught separately by a trained external coach, following a standardized lesson plan that ensured consistency across all classes. The plan covered six phases of debate and writing instruction and remained unchanged throughout the study. Control groups received regular German lessons from their teachers, focusing only on exam-related tasks and argument structures, without debate or SRSD training. To guarantee these conditions, agreements were made with the teachers.

Fidelity was maintained by having all interventions delivered by members of the research team who were trained to follow the jointly developed plan. No observations were possible in control groups. Data collection complied with the Ministry of Education’s requirement for anonymity. Students created individual five-character codes before the pre-test, known only to themselves, ensuring no identifying information was accessible to teachers or researchers. No further demographic data were collected.

2.3. Measures

All essays were written digitally by the participating students using Gorilla Experiment Builder (Build-v2025-09-05, UK) We chose this tool for this study because the independent survey design could be implemented free of charge without the need for extensive programming skills. In addition, the survey also used Gorilla Experiment Builder (in the following, briefly: Gorilla), which is compliant with the EU data protection directives (Anwyl-Irvine et al., 2020). References to Gorilla as a digital online platform were documented in numerous international studies between 2018 and 2023. An exchange with the Gorilla team revealed that no written material had been collected using this software prior to the start of this study. Gorilla converted the students’ written material into Excel format. From there, this material was transferred into suitable documents to forward it for the recording of text quality.

In general, text quality was measured at four different measurement points (see Table 1) during this longitudinal study using four different writing tasks in both the intervention groups (n = 237) and the control groups (n = 118). In all cases, the writing tasks are standardized writing exercises that have been used in this form in authentic examinations in German conducted at the end of grade 10. All these argumentative writing tasks had a similar structure: the students were required to write both pro and con arguments on a particular issue. One argument is given in a speech bubble for each position (see Appendix A.1, Appendix A.2, Appendix A.3 and Appendix A.4).

Moreover, text quality on all four measurement occasions was assessed according to the following levels: holistic and analytical text quality based on language pragmatics (e.g., content organization of the text and text structure). All the anonymized pro-and-con arguments (n = 1061) were assessed by two raters in a double-blind peer review based on IMOSS (“Integrated Model of School Writing”) encoding (Neumann, 2012; Neumann & Matthiesen, 2011).

The reliability of this coding method is at an appropriate level, with a Cronbach’s α = 0.90. Cohen’s Kappa measures inter-rater reliability; here, the value is in a satisfactory range (κ = 0.727) for text quality. However, it does not consider pieces of text that were declared as missing data, because they did not fulfil the task or were too short (less than two sentences). Regarding this material, however, it is unclear whether the participating students were not motivated or whether they did not have the skills required to write an argumentative text. Therefore, the IMOSS scale was expanded to 6 points for this study (1 = failed text; 6 = high-quality text). In the analysis, we both included and excluded off-task scores (1, failed text) to demonstrate their effect on the overall evaluation; the inclusion of scores equal to 1 was considered crucial to ensure the integrity of the dataset. Through this approach, a comprehensive view of the distribution and variability of the data was obtained. The inclusion of scores equal to 1 was also significant in terms of comparing the ranges of the students’ performance, especially for the struggling ones. Finally, a score of 1 provided baseline information on the effectiveness of the interventions, while the analysis carried out with scores of 1 excluded indicated the interventions’ efficacy in a more selective way that helped us interpret the data.

Previous SRSD studies used 6-point scales to measure text quality (E. K. Washburn et al., 2016; Chalk et al., 2005). From a psychological perspective, using this scale reduces central tendency bias, avoiding extreme rating categories. It aids in clearly representing raters’ judgments, thus helping produce richer data with potentially varied responses and allowing a more nuanced assessment of attitudes and perceptions.

The research team trained the two raters independently of each other in the use of the IMOSS coding procedure. As Rijlaarsdam et al. (2012) emphasized, independently training raters to use a reliable coding procedure is a prerequisite for obtaining a rating procedure that is as error-resistant as possible. Further criteria for the rating procedure were considered by defining benchmark writing pieces prior to the rating to allow higher inter-rater reliability, and they were defined by both raters coding the texts both holistically and analytically for each text using the same procedure. As a consequence of this rating procedure, all the texts were double-blind-rated. The inter-rater reliability for the 1061 texts in this study was also in a satisfactory range (κ = 0.828) for text quality.

SPSS 28 and R were used for descriptive statistics and inferential statistics, respectively. Multi-level models were used primarily for variance decomposition and obtaining exploratory insight, not for hierarchical dependency modeling. Despite the presence of hierarchical or nested structures in the data (for example, students nested within schools), the primary focus of this study was on conducting between-group comparisons rather than explicitly modeling multi-level dependence. The experiments were conducted in distinct groups whose values were not repeatedly measured within nested structures in a manner requiring multi-level modeling.

The analysis revealed that most of the variance in the outcome variable, text quality, was between participants and learning groups. The minimum variance was observed to be consistent among the schools. Notably, the total participant variance was 38%; 16% was attributed to classes, and only 6% was attributed to the schools, showing that schools could be excluded as a level if needed. In a comparison using only the learning-group levels, the model with three levels (with the participants as level 2 and class as level 3) explained significantly more of the variance. The context variable accounted for 36% of the total residual variance, while other individual differences, including those caused by the training measures, accounted for 64% of the error variance.

The primary objective of this study is to examine variations in the development of argumentative text quality across different intervention groups, school types, and intervention orders. The analyses of independent sample t-tests and univariate ANOVA were deemed appropriate because the data points—such as pre- and post-intervention scores—were considered independent observations within the specified groups.

Moreover, preliminary assessments indicated that the variance attributable to the nested structure (i.e., differences between schools) was relatively small and did not significantly influence the outcome variables. Consequently, the utilization of simplified parametric tests, such as t-tests and ANOVA, furnished a straightforward and statistically valid approach for the analysis of group differences and developmental changes while also maintaining analytical simplicity and interpretability. To address these research questions and test the hypotheses, a multi-level analytical approach was employed.

Additionally, a three-level linear mixed-effects model was used to analyze the data, utilizing R (Version 4.3.2; R Core Team, 2023), RStudio (Version 2024.09.1+394; Posit Software, PBC, 2024), and the lme4 package (Version 1.1-34; Bates et al., 2015). In this model, Time (t₁–t₄) was nested within Students, and Students were nested within Classes. The fixed effects included intervention group, time, school type, and their interactions. This approach accounted for the complexity of the data structure and provided insights into how text quality varied across different conditions.

Missing data (<5%) were addressed using full-information maximum likelihood, which allowed for the inclusion of all available data without significantly impacting the analysis.

An independent-samples t-test was conducted to compare the Mean Overall Impression (MOI) and Mean Language Pragmatics (MLP) scores between the two intervention sequences (debating–SRSD vs. SRSD–debating). This analysis specifically addressed RQ₃, which concerns the effectiveness of intervention order.

Pearson’s correlational analysis was performed to investigate the relationships between time point, school type, intervention group, and MOI and MLP scores, providing valuable information relevant to RQ₂, which concerns how school type influences writing performance.

Univariate ANOVA was conducted to examine the effects of time point, school type, and intervention groups on both MOI and MLP. This analysis helped evaluate the hypotheses related to RQ₁ and RQ₂, revealing significant differences in text quality across time and educational contexts.

The findings from these analyses are detailed in the following section, which presents the results concerning the participants’ argumentative text quality development throughout the study and the effectiveness of the various educational interventions.

3. Results

This section addresses the research questions (see Section 3) based on an investigation of the development of students’ argumentative text quality (RQ₁), the influence of different school types (RQ₂), and the effect of intervention order (RQ₃).

3.1. Descriptive Analysis

Table 2 presents the means and standard deviations for text-quality scores, specifically the Mean Overall Impression (MOI) and Mean Language Pragmatics (MLP), across four time points (t₁–t₄). The scores are disaggregated by intervention group—debate–writing, writing–debate, and control—as well as by school type (Lower Secondary School [LSS] and Higher Secondary School [HSS]). The values indicate that both experimental groups showed progressive improvements in writing quality over time, while the control group showed only minimal gains. The sample sizes (n) reflect the number of students contributing data within each condition and at each school level.

3.2. Correlational Analysis

School type exhibited a strong positive correlation with Mean Overall Impression (r = 0.26 **, p < 0.01) and Mean Language Pragmatics (r = 0.27 **, p < 0.01), demonstrating that the type of school has a strong influence on these two measures. There was no significant correlation between time point and either school type (r = −0.04) or intervention group (r = −0.02), demonstrating the ineffectiveness of these factors over the change in time points. However, there is a moderate positive correlation with Mean Overall Impression (r = 0.26 **, p < 0.01) and Mean Language Pragmatics (r = 0.25 **, p < 0.01). The scores of both tend to increase with the time factor. No significant correlations between school type and intervention group were found (r = −0.03).

Intervention group displayed small negative correlations with Mean Overall Impression (r = −0.05) and Mean Language Pragmatics (r = −0.04). The non-significance of both these variables shows that the assignments employed in the intervention group did not have a strong effect on performance for these measures. Mean Overall Impression exhibits a strong positive correlation with Mean Language Pragmatics (r = 0.94 **, p < 0.01), indicating that the performance for one measure is highly correlated with the other measure, and the strong correlation among the measures also supports the uni-dimensionality of the compound scores. This analysis was conducted to test RQ₂: how does school type affect students’ argumentative text quality under debating and SRSD intervention conditions?

3.3. Independent-Samples T-Test

We compared the two intervention groups (DW and WD) in terms of Mean Overall Impression (MOI) and Mean Language Pragmatics (MLP). Levene’s Test for Equality of Variances indicated that the assumption of equal variances was met (F = 2.66, p = 0.10), allowing us to proceed with the results with ‘Equal variances assumed’. There were no significant differences between the groups, confirming our hypothesis that intervention order would not influence the results. Similar mean scores were observed (e.g., MOI: DW = 3.20, WD = 3.21; p = 0.70).

The mean scores for Mean Language Pragmatics were equally comparable and indicated that intervention order did not significantly affect the outcomes. The analysis revealed that mean differences and effect sizes were minimal (e.g., Cohen’s d = 0.01, and Hedge’s g = 0.01), reinforcing the conclusion that the order of intervention does not significantly impact the results. This analysis was carried out as an extension of RQ₃, which concerned determining which order of the interventions (writing–debate or debate–writing) is more effective for developing argumentative writing skills.

3.4. Univariate ANOVA

The analysis was based on multiple observations made by the same participants, with the participants treated as the random effect to account for individual variability, enabling us to acquire different baseline scores and responses in order to increase the generalizability of the findings. The intervention groups (DW and WD) and control group are the fixed effects in the model because the influence of intervention groups was observed in terms of dependent variables. Similarly, the time points (t₁, t₂, t₃, and t₄) were also considered fixed effects, enhancing their influence on the outcome variables. Furthermore, the interaction effect between time point and intervention group was assessed.

The effects of time point, school type, and intervention group on Mean Overall Impression and Mean Language Pragmatics were examined using univariate ANOVA. Significant effects were observed for the time point factor (MOI: F(3,1037) = 70.66, p < 0.001; MLP: F(3,1037) = 71.59, p < 0.001), indicating an improvement in text quality over time. There was also a significant effect of school type (MOI: F(1,1037) = 97.37, p < 0.001; MLP: F(1,1037) = 108.55, p < 0.001), showing that the HSS students scored slightly higher than the LSS students. Furthermore, the results suggest that both instructional sequences are equally effective, as neither the intervention group nor the interaction terms (e.g., time point * intervention group) yielded significant results. The analysis was based on RQ₁ and RQ₂, analyzing how the participants’ argumentative text quality developed throughout this study under debating and SRSD intervention conditions, depending on whether they were in one of the intervention groups or the control group, and, similarly, how school type affects students’ argumentative-text quality development under debating and SRSD intervention conditions.

4. Discussion

This study focused on the effect of two combined approaches, namely, debating and Self-Regulated Strategy Development (SRSD), on the argumentative writing skills of German ninth-grade students. Additionally, the SRSD approach was adapted for German schools in the context of argumentative writing for students in grade 9. The findings of this study provide a foundation for the growing literature on oral and written argumentation strategies and enrich argumentation-teaching research through this investigation of a new approach combining debating and SRSD. The three research questions are answered below.

4.1. Discussion for RQ₁

The results derived from the descriptive statistics highlighted a steady increase in text quality across all groups, with significant improvements observed during the post-test phase and a slight decrease during noted the maintenance phase. This pattern is also consistent with observations made in other empirical field studies, where retention effects seemed to diminish over time (Felton et al., 2015). The SRSD approach combined with debating seemed to prove its efficacy in nurturing structured and self-regulated writing processes among students with limited knowledge of pro-and-con writing. The results correspond to the conclusions drawn by Graham et al. (2023) and Sun et al. (2022), who mentioned the impacts of this approach on writing and self-efficacy. Notably, students from a lower academic track benefited more from the SRSD instructions plus debating than their higher-academic-track counterparts, emphasizing the value of scaffolding for students with limited exposure to complex writing tasks (Graham & Harris, 2017; Mason et al., 2017; E. Washburn et al., 2016; Graham & Perin, 2007). In addition, this study recapitulates the results of previous studies, which showed the positive effect of SRSD intervention on argumentative writing skills (Mason et al., 2017, 2013; De La Paz, 2005).

Moreover, debating can serve as a motivational pathway with respect to students’ engagement in argumentative writing. As mentioned in previous studies (Malloy et al., 2020; Majidi et al., 2021), oral debates can facilitate critical thinking and the exchange of knowledge and perspectives and increase the coherence of arguments. The challenges of argumentative writing described in Section 1 can be addressed through debating, as students must consider not only their own position but also that of their opponents when preparing for the debate. In the debate itself, they can respond directly to the arguments of the opposing side without having to imagine the interlocutor. The linking of pro-and-con arguments is initially practiced orally in a debate via the students first explicitly referring to the opposing sides’ arguments. This can facilitate transfer to written pro-and-con argumentation, as at least two perspectives are linked in the debate on an issue, which can be chosen by the students.

Furthermore, the defined structured format of debate according to the Jugend debattiert model seems to facilitate the transfer of declarative knowledge about argument structure into written tasks. This assumption is emphasized by the results at t₂ for debating groups, showing better text quality without SRSD writing lessons. The structure of a debate, consisting of an opening speech, free debate, and a closing speech, is similar to the structure of an argumentative text, with an introduction, a main part, and conclusions. The sub-elements have similar functions: introducing the topic, weighing the pro and con arguments, and summarizing the arguments. This transfer can be illustrated for students by formulating a debate in an argumentative text, or vice versa (see RQ₃).

Thesis₁: The data obtained in this study indicate that written argumentation training based on the SRSD approach and debating training improves the ability of ninth graders to write comprehensive pro-and-con arguments (thus confirming H₁).

4.2. Discussion for RQ₂

First, Pearson correlational analysis revealed that school type has a strong positive correlation with the Mean Overall Impression and Mean Language Pragmatics, demonstrating that regardless of whether students attend an LSS or an HSS, they will exhibit different performance for both of these measures. Additionally, the significant difference between HSS and LSS students regarding text quality is related to the aspects of the curriculum. While HSS students in Germany begin writing pro-and-con arguments in the eighth grade, LSS students are not required to write arguments until the ninth grade. This means the HSS students in this study had more prior knowledge of pro-and-con argumentation. This is particularly important because the final examinations at the end of grade 10 take place just one year after LSS students are taught how to write pro-and-con arguments. Pro-and-con argumentation is a compulsory part of this examination. Based on this study, starting to write arguments earlier on in LSS can also be recommended, as it provides students with as many opportunities as possible to work on their writing skills.

Secondly, this study confirms the results of previous intervention studies that have indicated that students with little prior knowledge of argumentative writing (as in those from German lower-academic-track schools) especially benefit from a series of lessons based on the SRSD approach.

Thesis₂: Our findings show that the students from higher-academic-track schools produced better text during the study than the students from lower-academic-track schools because of the curricula regarding argumentative writing. However, the LSS intervention group’s improvements show the potential of combining debating and SRSD interventions (thus, H₂ can be confirmed).

4.3. Discussion for RQ₃

This study clearly shows the benefits of an approach that combines the promotion of the development of oral and written argumentation skills and SRSD on improving the argumentative writing skills of ninth-grade students. This study assumes that incorporating the socio-affective aspects of oral debating could influence the motivation of disadvantaged learners and support inclusive education. Further intervention studies are needed to reveal the influence of debate training on students’ writing skills. This raises the question of the extent to which oral argumentation supports the writing of argumentative texts or, in other words, how the transfer of competencies improves writing skills. The statistical analysis suggested that there were improvements in the students’ argumentative writing across all the intervention groups, with significant improvements before and after the test. Moreover, there was no significant difference in the results between the debate–writing and writing–debate groups; this result also aligns with the findings of previous research emphasizing the independent efficacy of both debating and SRSD interventions in improving argumentative writing (Majidi et al., 2021; Mason et al., 2017).

The results show that students can benefit from either order of interventions, allowing flexibility in instructional design. This adaptability is critical for inclusive classroom settings, where students’ motivation and levels of skill differ significantly (Graham et al., 2023). Whether the promotion of oral argumentation skills through debate training occurs before or after an SRSD intervention does not seem to play a decisive role. This finding is contrary to the findings reported in previous studies, which recommend promoting oral argumentation before writing (Majidi et al., 2021; Malloy et al., 2020; Reznitskaya et al., 2001).

Thesis₃: Based on the results, we assume that intervention order (writing–debate or debate–writing) does not play a role in developing writing skills (thus, H₃ should be rejected).

4.4. Synthesis and Theoretical Significance

This study confirms that dialogical argumentation—through debate or SRSD—effectively supports argumentative writing. Instructional gains were evident compared to the control group, with particularly strong effects among LSS students, who likely benefited from scaffolding that had been absent in their curriculum. Interestingly, the order of interventions did not influence outcomes, indicating that schools can flexibly sequence them without compromising effectiveness.

The findings build on theories of self-regulated learning, emphasizing meta-cognitive techniques such as goal-setting, monitoring, and strategy evaluation (Zimmerman, 2002). SRSD, grounded in these principles, enables students to plan, monitor, and revise their work more effectively (Paris & Paris, 2003). Integrating meta-cognitive training into argumentation instruction increases awareness of writing processes, deepens engagement, and fosters autonomy. Further research should examine how specific meta-cognitive prompts support the transfer of argument structures across subjects.

The quasi-experimental design provided practical classroom insight but limited full control. Novelty effects from researcher involvement cannot be ruled out, and isolating variables in authentic classrooms remains challenging. Nevertheless, the quantitative analysis revealed meaningful group differences, while future research could combine this with qualitative approaches for richer understanding.

Argumentative competence is shaped by both cognitive and affective factors. Debating not only enhances reasoning and language skills but also builds motivation, self-efficacy, and confidence (Bandura, 1997; Pajares, 2003). The dialogic nature of oral argumentation supports the transfer into writing and motivates students by framing argumentation as persuasion. Instruction that fosters both cognitive and affective growth is therefore essential.

Cultural and curricular contexts also matter. In Germany, argumentative writing is formally taught at specific grade levels, reflecting cultural influences (Mason et al., 2017). Adapting pedagogical models like SRSD and debating to different educational traditions is vital for global relevance. Expanding international research would strengthen cross-cultural comparisons and support the development of broadly applicable models (Kuhn & Crowell, 2011; Achour et al., 2020).

Despite evidence of effectiveness, practical challenges remain. Teachers face time constraints, limited training, and resource shortages (Graham et al., 2022). Greater emphasis on scaffolding techniques, targeted assessments, and the integration of digital platforms could help. Formative assessment and peer feedback are especially important for sustaining improvement and promoting reflective classroom cultures (Black & William, 2009).

Within Germany, more attention must be given to the curricular role of argumentative writing, particularly in secondary schools where pro-and-con texts are required. Instructional strategies like SRSD and debating need stronger integration to ensure all students acquire these skills (De La Paz, 2005; Graham & Harris, 2017).

Although many studies report positive impacts, there is limited understanding of the mechanisms underlying the effectiveness of SRSD and debating (Graham et al., 2023). Both methods offer differentiated support, such as scaffolds for students who need additional assistance, opportunities for advanced learners, and group-based debates that may help reduce cognitive load. Further research should examine these mechanisms in greater detail to improve models and optimize outcomes.

Collectively, these findings contribute to existing argumentation models by indicating that cognitive (strategy-based) and socio-dialogic (debate-based) approaches can complement each other when implemented together in inclusive classroom settings. The results indicate not only the potential effectiveness of each method but also the possible benefits of their integration in addressing disparities in argumentative literacy. By embedding strategy instruction within a context that has social relevance, the results of this study provide insights into ways that argumentation skills might be developed in secondary education and offer a model that could be adapted for different cultural and curricular environments.

5. Conclusions

5.1. Summary of the Study

This study investigated the effects of combining Self-Regulated Strategy Development (SRSD) writing instruction with structured debate formats on the argumentative writing skills of ninth-grade students in German secondary schools. Using a quasi-experimental design with six intervention and six control classes across both higher and lower academic tracks, the study examined changes in text quality across four measurement points. The findings showed that both SRSD and debating improved students’ argumentative writing, with no significant differences in effectiveness between intervention orders. Importantly, students from lower-track schools benefited disproportionately, suggesting that the combined approach has particular value for addressing curricular inequalities in Year 9.

5.2. Pedagogical Implications

The main goal of this study was to investigate how the quality of argumentative texts written by ninth graders attending six Higher and Lower Secondary Schools in Germany could be assessed and promoted via two combined debating and SRSD interventions. Considering the results, the following conclusions can be drawn for future empirical research and school practice.

Multimodal approach to promote debating and argumentative writing. This study demonstrates how debating and SRSD approach can be effectively linked within an intervention framework. Descriptive statistics show the potential of both debating and SRSD for promoting writing, with debating having a particularly strong impact at the second measurement point (t₂). This underscores the value of integrating debating into the curriculum as preparation for written tasks. By selecting topics, researching, evaluating sources, organizing evidence, and the change of verbal interaction and deepening through writing students develop multimodal skills that strengthen both oral and written argumentation.

Flexibility in learning paths for all-in-one class. This study shows that the sequence of educational measures—whether starting with debating or writing—does not significantly affect writing development. This flexibility allows teachers and students to choose learning paths based on individual motivations and abilities. For instance, students with prior debating experience (e.g., from academically oriented schools) may benefit from beginning with debates, while others may prefer writing first. A recommended lesson structure alternates 20-min silent and action-based units, enabling parallel debating and writing in different groups. Such adaptability fosters inclusion, engagement, and differentiated learning. However, as this was a pilot study, further controlled research is needed. Teachers are encouraged to let students select their paths and adapt instruction to diverse needs, as shown in the SRSD and Jugend debattiert phases (see Table 1). The study is particularly relevant because it presents concrete, empirically derived recommendations for teachers on how to promote written argumentation among students in different tracks.

Challenges of school system inclusivity. This study found that school type did not affect intervention success but did influence students’ preparedness for argumentative writing. HSS students were generally better equipped than LSS students, despite limited practice opportunities after grade 10. These findings highlight the need for curricular reforms to ensure equitable preparation and align with previous research (e.g., Giera, 2020).

Innovative measurement techniques. For the first time, Gorilla Experiment Builder (Build-v2025-09-05, UK) was used to evaluate the writing process and assess students’ attitudes toward writing. In collaboration with the software team, the design was adapted for use in schools, enabling digital measurement of reading, planning, and writing times. This integration provides insights into the cognitive and emotional dimensions of writing and allows teachers to deliver targeted, individualized support that would be impossible in large classes.

5.3. Limitations

This pilot was one of the first quasi-experimental long-term study in Germany to combine debating and written argumentation, and one of the first to adapt the SRSD approach for pro-and-con argumentation. These contributions must be balanced against several limitations.

The quasi-experimental design prevented full randomization; classes were assigned to intervention or control groups, so class-related effects cannot be ruled out. Although participants were randomly distributed across DW and WD panels, the involvement of external coaches and researchers may have introduced a novelty effect. Teacher-led studies with longer follow-ups are needed to test sustainability. Data collection by teachers, the small sample size, and the absence of control-group observations further limit generalizability. Some teachers also conducted class tests despite agreements not to, which may have influenced outcomes.

The analysis focused on Mean Overall Impression and Mean Language Pragmatics, excluding language systematics (grammar, spelling, punctuation), as it was not central to the intervention. Only 44 students could be tracked across all four time points, restricting individualized progress diagnostics, though groups remained stable.

Finally, the study was conducted during the COVID-19 pandemic, requiring occasional online lessons. Future research should use larger samples, teacher implementation, and qualitative approaches to deepen insights.

5.4. Future Research

With respect to the presented limitations, future research should integrate quantitative outcomes with qualitative insights from observations, teacher reflections, and student interviews to clarify underlying mechanisms. Longitudinal studies could show whether benefits persist once integrated into regular practice. Despite limitations, the study highlights the potential of multimodal approaches linking debating and written argumentation by the SRSD approach. Engaging students in debating, reflecting, and writing fosters competencies essential for democratic participation, critical thinking, and lifelong learning. Bridging oral and written argumentation through debate and SRSD can promote more equitable opportunities for all learners to develop strong writerly voices in inclusive classrooms.