In 2019, in the German state of Northrhine-Westphalia introduced a new curriculum for the subject of “Economics & Politics” at the lower secondary level which replaced that of the former subject “Politics/Economics”. Concurrent with this curriculum reform, the state government reestablished the lower secondary level with nine terms as the standard in lieu of eight terms, which had existed previously. The additional school year was predominantly utilized so as to emphasize economic education within the new subject. As the reform was highly controversial among the general public and among teachers, questions arose regarding the acceptance of the new curriculum by those who had to work with the new curriculum: teachers. The presumption was that acceptance would vary, given the diversity of teachers who teach the subject. Consequently, the initial step was to identify various types of teachers for this particular subject, which makes up the focus of this paper. The study of the acceptance of the reform is the second step.
It is essential to realize that subjects and “subject matter is mediated by teachers and students in face-to-face classroom practice into an ‘enacted curriculum’” (Deng, 2008, p. 67; Hericks, Körber, 2007) and thus the curriculum of a subjects a selection “of cultural canon and memory, scientific truth, method and knowledge, and aesthetic artifacts and performances”.
The theoretical literature argues that teachers set their preferred didactic teaching style in accordance to their individual action-guiding notions as pointed out by Hericks and Körber (2007). Our key interest is to test this framework of teacher behavior by conducting a survey among high school teachers and extend their framework to allow for subject-didactical heterogeneity. We confronted teachers with a questionnaire asking about their preferred 1. subject-didactical teaching approaches, 2. learning tasks and 3. subdisciplines in social sciences. Assuming a large degree of heterogeneity among teachers, we developed an empirical method how to identify different types of teachers conditional on their differences. This empirical methodology can be applied to other subjects such as geography and languages.
Of particular interest is thus the performance by teachers, in other words action-guiding notions and pedagogic orientations of teachers, which are part of face-to-face classroom practice (Hericks, Körber, 2007). Consequently, this study primarily focuses on said notions of the educators with an emphasis on typical notions, e.g. we here propose a subject-didactical typology of social science teachers. Furthermore, we posit that within a subject, there is not just one disciplinary culture, but multiple co-existing sub-disciplinary cultures (ibid., p. 36), especially in a subject like social sciences combining multiple academic sub-disciplines (Löfström, Grammes, 2020). This results in a typology of social science teachers who in turn differ in their teaching. The methodology explored in this study can be applied to other subjects with several sub-disciplines. The results can be used to study the acceptance of the reform as well as student output as a function of teacher type. We therefore offer a subject-didactical addition to the professional development (PD) literature suggesting, that there are several content-specific pathways to professionalism within given subjects.
Characteristics of teachers are crucial for understanding and developing instructional practice in educational science. Several strands of literature have contributed to this field. One subfield focuses on teaching expertise (Raduan, Na, 2020; Anderson, Taner, 2023; for social science teachers cf. Crocco, Livingston, 2017). Related to this strand is the domain of teacher professionalism research (Darling-Hammond et al., 2017; Edovald, Nevill, 2021, Lynch et al. 2019, Stronge et al. 2011) that attempts to link individual teacher attributes and types of teachers to instructional quality and ultimately teaching effectiveness (Scheerens, 2007, Scheerens, Blömeke, 2016, Scheerens 2017, Senden et al., 2021).
The empirical strand of this literature on teacher effectiveness (Hattie, 2012, Gates Foundation, 2010) rests on the assumption, that whatever teachers do in class is “the major source of controllable variance” (Hattie, 2012, p. 25.). This is why studies have focused on teacher characteristics as main explanatory sources of student outcomes. The main assumption of this strand of literature is that successful teacher professionalization takes place with a content focus, which “generally treats discipline-specific curricula” (Darling-Hammond et al. 2017, p. 5).
One missing source of variance that has not been yet controlled for and has, so far, not been introduced into the theoretical literature is teachers’ varying subject didactical perspectives. While subjects come with a specific perspective of the world, it is fair to assume that these perspectives shape how and what teachers present students in terms of teaching material and tasks and what students in turn do in class. A typology of teachers rooted in subject matter didactics is thus significant because instruction always takes place within subjects – hence the focus of professional development research on disciplines – and thus learning by teachers can manifest itself in the acquisition of both general pedagogical, and subject-specific competencies, consisting of content knowledge and pedagogical content knowledge (Shulman, 1986).
A subject-didactical typology of teachers builds on the findings of existing literature for the following reasons:
First, one key assumption of input-process-output models of instruction, that are the dominant models to assess instructional outcomes, holds that that different teachers offer different types of input (i.e. instruction) within a subject (Creemers, 1994, Scheerens, 2007, Helmke, 2021). These models hypothezise that teachers offer input in line with process characteristics as well as personal characteristics. The input in turn is interpretatively used by students, resulting in different individual interpretations and thus different outputs: “The instruction itself is influenced by teacher attributes, especially their subject-matter and didactical expertise, as well as the attitudes, orientations, and standards of the teacher” (Brühwiler, Helmke & Schrader, 2017, our translation). If this also withstands closer empirical scrutiny, we should treat teachers as a non-uniform group with different perspectives on the same subject. This, too, accounts for the need of a subject-didactical typology that allows to control for this yet unobserved heterogeneity in the above-mentioned literature on teacher professionalization.
Second and related to the local settings of teaching in the state of Northrhine-Westphalia, the school subject of social sciences integrates three sub disciplines namely political science, sociology, and economics. These three disciplines follow different academic perspectives and methodological approaches. Therefore, it is expected that this disciplinary triad is also reflected within the teaching community with teachers choosing different key aspects in their teaching.
Third, it can be assumed that depending on differences among ideal notions of instruction, different educational approaches are chosen. In recent years, evidence-based approaches and data-driven methods have gained in importance in economics, a method that has always played a major role in sociology, while quantitative-empirical approaches seemed less relevant in political science. Thus, we hypothesize that these different approaches can be surveyed not only through self-reports about a teacher’s instruction but also through preferences for specific task types reflecting the different academic methodology of sociology and economics, and political science. These can, in turn, be used for a subject didactical justified typology of teachers which we will look into in the next paragraph.
Albeit these reasons that conjointly suggest the development of a typology that allows us to respect the didactic heterogeneity in professionalization literature, we are also aware that such an endeavor comes with noise and possible misspecification: Two main reasons are as follows:
First, we want to skim expectation of such a typology since we find a multiplicity of factors that produce noise in this process. Second, we are not the first who suggest that such a typology is a meaningful addition for the professionalization of teachers. Schiller (2009) identifies four distinct teacher types (Schiller, 2009) by analyzing teacher preferences on tasks (ibid., p. 251). With our redefined modeling and approach, we tried to further their contribution with a more rigid methodology.
In the course of the competency-oriented turn in the field of didactics learning tasks become more important. Competencies are conceptualized as “context-specific cognitive dispositions that are acquired by learning and needed to successfully cope with certain situations or tasks in specific domains.” (Klieme et al., 2008, p. 9, Klieme et al., 2009). This implies that during instruction, students are placed in subject-specific situations in which they must address subject-specific requirements using subject-specific methods. These situations, requirements, and methods are presented to students by way of teaching materials and learning tasks. Tasks are often conceptualized as “learning opportunities” (Scheerens 2017, p. 43) Oelkers and Reusser accordingly demand that “good subject-specific learning tasks materialize those components of knowledge and ability, trigger those cognitive and work processes, and activate those analytical and synthetic figures of problem-solving, arguing, considering, and interpreting that are central to a particular subject and constitute its intellectual culture.” (Oelkers, Reusser, 2008, p. 408, our translation). If used appropriately, effective learning tasks allow for the “re-transformation of competency standards into meaningful and sustainable learning activities” by way of a high degree of cognitive activation (ibid, p. 410, Blömeke et al., 2022). Thus, textbooks which contain material and tasks can be considered the backbone of instruction of student-centered learning environments (Hoidn, Reusser, 2020) which predominate today. Learning tasks determine the progression with which students approach the learning objective and, ideally, ultimately accomplish the respective learning objective. This is reflected, for example, in natural science didactics with the backbone model of planning as well as in the task-based learning approach in foreign language didactics (Willis 2017). In the didactics of social sciences, the role of tasks has so far only been discussed in the context of problem-oriented teaching (as a precursor to competence-oriented teaching, see Roth, 1970, Klieme et al., 2008) and discovery-based learning (Bruner 1981). For mathematics instruction, the international TIMSS study found that “tasks administered play a key role in promoting students’ […] learning—and that tasks are thus indeed suitable indicators of cognitive activation in the classroom“ (Neubrand et al., 2013, p. 125). While the focus of TIMSS and the COACTIV study (Kunter, Voss, 2011) was on the extent of cognitive activation, considering an integrated subject like social sciences with its three sub disciplines, one can also assume that, depending on the teacher, foci in content and methodological approaches manifest themselves through the selection of corresponding tasks. In input-process-output-frameworks, the preference for certain task types can be conceptualized as a characteristic of teachers that influences the input side of teaching. Accordingly, it is to be expected that teachers, based on their task preference, provide different instruction, and generate different outcomes for different students. This is not to be valued as a negative statement in the sense that it is about more or less, but rather orthogonal to that about a different kinds of teaching, which is initially neither good nor bad in terms of outcomes but will generate different outcomes for different students.
Tasks in social science instruction can target different teaching material. In general, a distinction can be made between text-based teaching material and data-based teaching material in different forms of presentation (see also Engartner, Hedtke, Zurstrassen, 2020, p. 136f., Goldschmidt, Kron & Rehm, 2024, p. 10). This distinction reflects an approach to instruction that is based on theories versus one that is more empirically based. Learning tasks in social science instruction can also aim at different performance levels. We assume that tasks at the level of performance level I (reproduction of knowledge) are used by all teachers. Beyond that, a focus on either performance level II (application of knowledge) or III (judgement) is conceivable. Lastly, within performance level III, considering Reich's distinction regarding constructivist activities in instruction, a focus is conceivable either in the area of deconstructivist tasks or of constructive tasks (Reich, 1996). Constructive tasks aim at the action-productive invention of reality; deconstructive tasks, on the other hand, focus on the critique of the reconstructed reality in terms of alternatives. In social science instruction, these two perspectives are prototypically found in the task operators “design”, “develop”, and “design” on the one hand, and “discuss”, “debate”, and “problematize” on the other hand (see MSW NRW, no date).
In research on ideal concepts of instruction, there are various paradigmatic approaches. Over the last two decades, Hilbert Meyer's ten characteristics of good instruction have probably been most influential in Germany (Meyer, 2016) with the seven Cs (Gates Foundation, 2010) being the most prominent in the English-speaking hemisphere. However, these and others relate to instruction in general and have limited significance at the level of subject-specific instruction. Subject-didactic ideal notions of instruction were described in the subject of social sciences mainly in the form of instructional guidelines within curricula with that of 1999 being the most elaborate. They were derived from the social science didactics at the time and are still binding. These instructional guidelines are: situational orientation, actuality, multiperspectivity of teaching, student orientation, education to democracy, reference to the real world, as well as action and product orientation (MSWWF, 1999, p. 8). The guideline of democratic education refers primarily to a political didactic goal. Student orientation aims in a similar direction, in the sense that the interests and desires of the students regarding the design of the instruction should be taken into account (MSWWF, 1999, p. 8). Thus, it is a non-subject-didactic guideline for instruction. However, in the German context of our sample, student orientation is often understood differently, for example, in the sense of a constructivist, student-centered rather than teacher-centered instruction with a transmissive impetus (Hartinger, Kleickmann, Hawelka, 2006, Welniak, 2017). Even in this case, it would still fall under the general didactic design guidelines. In the survey we did not disclose to teachers which interpretation we support but treated both as general didactic guidelines. Although the design guidelines are equally binding in their implementation, it is plausible that teachers prioritize them differently in their instruction (see above), even if this has not yet been researched. In addition, given the tripartite division of the school subject's sub-disciplines (political science, sociology, and economics), we study to what extent teachers identify with one or more of these sub-disciplines and consequently align their teaching. Thus, we are inquiring about teachers’ statements about the relative importance of competence orientation in political science, in sociology, and in economics, respectively, in their lessons.
The survey was fielded during the implementation events of the new curriculum for Economics & Politics for the lower secondary level of high-schools in the state of Northrhine-Westphalia during fall 2019. We were able to collect survey responses by department chairs of almost every school in the Westphalia region who were present during these events of the district governments in the three Westphalian administrative districts namely Arnsberg, Detmold, and Münster. In the Arnsberg administrative district, three implementation events were carried out and visited by our team. One each was attended in the Münster and Detmold administrative districts. Out of a maximum of 315 spokespeople of the social sciences departments in 315 secondary schools, data was retrieved from 240 individuals, which corresponds to a response rate of 83.3%. The difference can be explained by non-attendance at the implementation events and non-completion of the questionnaires. With 4,088 teachers of Economics & Politics at the lower secondary level in Northrine-Westphalia in the school year, it makes up 5.87% of all teachers in the field.
In comparison to the average teaching staff in Northrhine-Westphalia in the year of the survey, the sample is more female and younger (cf. tables 1 and 2). As teachers typically teach an additional subject besides Economics and Politics, we also assessed which additional subject they teach: German, English, and History were the most prevalent subjects, which is equivalent to the rest of the state (see MSB NRW, 2019, p. 49). These three subjects all have in common that text production and text analysis occupy a broad space in the instructions. We classified these subjects as “text subjects”. In contrast, natural sciences subjects as additional subjects alongside social sciences, are underrepresented (cf. table 3). Other sociodemographic data besides gender, age range, and additional subjects were not collected for data protection reasons, as this would have made it possible to assign an otherwise anonymized answer to a specific head of department.
Nine statements about teachers’ teaching, which originate from the 1999 curriculum for social sciences, were used as variables. The statements are: importance of 1. Up-to-datedness of topics, 2. multiple perspectives, 3. student orientation, 4. relevance of instruction to students’ life, 5. action orientation, and 6. democracy education. Regarding the subject focus of the lessons, questions were asked about the extent to which 7. competence orientation in the field of economics, 8. in sociology, and 9. in political science are central to teachers’ teaching.
For task preference, ten typical tasks from the central high school graduation exam (German equivalent of A-Level exams) were constructed, each showing a manifestation in the three-dimensional space regarding performance level II vs. performance level III, text-based vs. data-based activities, and constructive vs. deconstructive activities. All these tasks were related to macroeconomics to avoid revealing teachers’ different content preferences.
To reduce dimensions, a principal component analysis was conducted. Based on the correlation matrix of the 18 potential variables, seven were excluded from further analysis due to insufficient correlations (<0.3). Thus, eleven variables remained for further analysis. Due to the Likert scaling and therefore non-metrically scaled variables, polychoric correlations were used. With a KMO of 0.816, an anti-image consistently above 0.5, and a Bartlett’s r of 0.000, the solution is usable. The scree plot and the corresponding eigenvalue data suggest a solution with three factors (three factors have eigenvalues < 1), which explain 60% of the variance (cf. table 4).
The three factors were used in a cluster analysis to typify the teachers. After excluding outliers (checking for maxima and minima of Likert-scaled data to exclude wrongful inputs, listwise deletion), 213 cases remained in the sample. Since there is no correlation of the variables gender, age, and text subject these three variables were not used in the cluster analysis. To further validate the results, correlations of text subject, gender, and age with the three factors were calculated, which showed no significant correlation with any of the three factors. We chose complete linkage as the clustering method because this method produces relatively small but homogeneous groups. The sum of squared errors and interpretability suggest a solution with five clusters.
The principal component analysis with eleven variables related to instruction design and task preference yielded three factors which explain 60% of the variance. The three factors used are “subject didactical orientation”, “subject competency orientation”, and “data affinity”. The factor “subject didactical orientation” includes the variables student orientation, action orientation, democracy education, and relevance of instruction to students’ lives. Hence it emphasizes the role of student-centered teaching methods and education of values. The factor “subject competency orientation” covers the variables competencies in politics, in sociology, and in economics. Teachers who score high on this factor emphasize students’ expertise in the three sub-disciplines. Regarding this factor it is important to note at this point that there is no factor that focuses on a single sub-discipline. The self-assessment of teachers’ subject orientation therefore refers equally to the three sub-disciplines within the sample. Hence, in terms of self-perception there are no distinct teachers for politics, sociology, or economics, but they primarily consider themselves as social sciences teachers. The third factor, “text affinity”, comprises the two variables “determining” and “discussing” which represent the affinity for different types of learning tasks. High values of the third factor mean a preference for a deconstructive text-oriented task on performance level III compared to the sample mean (factor loading of −0.7056) and at the same time an aversion to a task that is constructive, data-based, but as well located at performance level III (factor loading of 0.6913).
For the cluster analysis, the three factors were used. The analysis yields five clusters that we refer to as “educators towards character formation” (30.0% of the sample), “subject experts” (10.7%), “infographics driven teachers” (5.6%), “traditional SoSci teachers” (27.5%), and “data driven teachers” (17.6%). The chart displays the respective average values of the three factors within the five clusters (cf. figure 1). The “educators towards character formation” are characterized by a relatively low emphasis on subject competency orientation and data affinity (equivalent to a preference for text-analyzing tasks). At the same time, they have a moderately high emphasis on the didactic-educational component, meaning an orientation towards democratic education and action-orientation among other student-centered teaching methods which ultimately aim at character formation. The group of “subject experts” has a relatively high emphasis on subject competency orientation and a relatively low emphasis on the other two factors. Thus, a focus on subject competency goes hand in hand with an aversion to data (and therefore a preference for text-based tasks) and a low emphasis of character formation. We can imagine a matter of fact teacher with little emphasis on different teaching methods on the one hand, meaning on the other hand regular usage of text analysis in classes as defining trait. “Infographics driven teachers” are characterized by a moderately high affinity for data-based tasks but score comparatively lowly on the other two factors, therefore neither seeing subject competencies nor character formation and different teaching methods component as central to their instruction. We can imagine a teacher who regularly uses data representation in classes, but has no high affinity to either of the subjects and methodological variety. The “traditional SoSci teachers” view themselves as a mix of experts and educators, but with a moderately high affinity to text-analysis tasks. Finally, the “data driven teachers” see data analysis combined with a subject competency orientation as central to their instruction (the polarization of text affinity is reversed here for practical reasons, i.e., data affinity is depicted). The subject didactical component is weakly emphasized among them compared to the rest of the sample. “Data driven teachers” are significantly more data-driven than the “educators” and the “subject experts”, and the “infographics driven teachers” are significantly less subject competency-oriented than the “experts”, the “SociSci”, and the “data-driven teachers”.
Although we nearly surveyed all spokespersons of social science departments in Westphalia, we can only indirectly infer the characteristics of the entire social science departments of upper secondary schools from these persons, as they are not representative of the department staff as a whole. In terms of age, the focus is on those between 30–40 years old, who probably perceive the role of the department head as a stepping stone in their career. Additionally, within the sample, women are underrepresented.
Furthermore, it remains unclear to what extent the self-reports regarding their teaching methods are valid. However, in the absence of the possibility for large-scale classroom observation, they provide an initial approximation for a subject-didactically justified, criterion-based understanding of instruction for the purpose of teacher typology. Statements about notions of ideal instruction can ultimately only be assessed by way of self reports.
In a further study we will examine if and in how far the five types differ with respect to the acceptance of the new curriculum. Based on their different subject-didactical approaches we can hypothezise, that data-driven teachers’ acceptance for the new economic parts of the curriculum is higher than the experts’ and the traditionalists’ as the latter have a preference for established, text-oriented teaching approaches which are not pre-dominant in economic education.
If and to what extent the five types of social science teachers can be found in other states and countries may be subject to further research. It is assumed that (slightly) different subjects result in different subject cultures which perpetuate by way of self-selection processes of teachers (Goldschmidt et al., 2024, p. 38). Social sciences, comprised of the aforementioned three sub disciplines primarily exist in the north-western German states, and in the form of citizenship education in many western countries (Löfström, Grammes, 2020) with distinct study programmes for prospective teachers. It remains unknown if the teacher types can be reproduced in these subjects, or if their subject culture is so distinct that there are completely different types. We are however confident to provide a typology that is easy to test and welcome to be replicated in other jurisdictions.
Future research may address the question, if different types of teachers lead to different learning outcomes by students, e.g. we can hypothesize, that “data-driven teachers” bring about better learning outcomes in terms of data competencies, while “traditional SoSci teachers” bring about better text-related competencies. This does not imply that one specific type causes “better” results. Moreover, different types may cause different results on the level of students. According to the input-process-output paradigm students choose and use lessons differently such that their preferences co-determines the learning outcome. Putting the typology in the center of this endogenous process might help to reduce complexity in uncovering the causes of learning.
Our study contributes to the literature with implications for other subject matter didactics independent of subjects and methodology as follows: First, our preference for task typology that allows for the analysis of preferences on reconstructive, constructive and deconstructive tasks can be transferred to many subjects. Such a subject-specific typology of teachers could be based, among other things, on a preference for this type of learning task. Second, preferences for different sub-disciplines of and their specific approaches manifesting in preferences for different tasks could be studied in other subjects, too. E.g. in geography, there has been a shift from physical geography to human geography (Unwin, 2013; Kesteloot, Bagnoli, 2021, Mattisek, Sakdapolrak, 2016). Human geography as a sub-discipline is conceptualized as a social science with the assumption, that classical approaches of physical geography could not explain questions of under-development, poverty, trade, etc. with physical features like natural resources and infrastructure alone. A preference for one of the two sub-disciplines might be reflected in teachers' different approaches to lessons in terms of contents and task preferences. In general science as an integrated subject in elementary school, humanities, social sciences and natural sciences are integrated into one subject (Smith Crocco, Livingston, 2017; Michalik, Murmann, 2007, p. 101ff). A type-specific prioritisation of content could be investigated here as well. Generally, competences in language teaching can be divided into listening, writing, reading and speaking skills (Burns, Siegel, 2018). So language teachers might have a preference in their daily practice towards one or more of these four competences which might reflect their preference for tasks, which is especially important in the light of the task-based learning paradigm in language learning (Willis 2021).
At last and with regard to professional development research, we posit that there might be several pathways to professionalism in subjects integrating several disciplines, depending on content-specific preferences of teachers.