Promoting Teachers’ Ability in Classroom Discussions About Controversial Issues Master Thesis by Olmo van der Mast
University of Amsterdam
Student: Olmo van der Mast Student number: 10686509 Supervisor: Jaap Schuitema Date: August 25, 2019
Abstract
In the present study, using an in-service teacher training, we sought to enhance secondary (vocational) schoolteachers’ ability to facilitate classroom discussions about controversial issues that promote students’ citizenship competences. First, using
questionnaires (n= 89) we examined whether the teacher training promoted teachers’ ability in facilitating classroom discussions about controversial issues that meet three important criteria (i.e., open and safe classroom climate, co-construction, critical thinking). Second, using interviews with teachers who participated in the teacher training (n= 6), we investigated if and how the training improved teachers’ practices in classroom discussions about controversial issues. We found some evidence for the effectiveness of an in-service teacher training in which teachers learned about and reflected on three important criteria of classroom
discussions (i.e., open and safe classroom climate, co-construction, and critical thinking) and different teacher roles. Yet, teachers still experienced several challenges during classroom discussions despite participation in the training. Perceived challenges/difficulties were: to maintain a safe discussion climate, in particular when the discussed topic may be emotionally charged for one or more students, when a controversial topic occurs spontaneously, or when extreme views turn up; to introduce relevant alternative perspectives to create a critical
discussion; or, to create an open discussion with a large group of students and/or limited time. Future studies could further investigate how teachers could deal with these challenges. In addition, future research could investigate to what extent teachers are able to engage their students in high-quality discussions that promote citizenship goals, and how this relates to teachers’ perceived ability in teaching controversial issues.
Keywords: classroom discussions, controversial issues, democratic citizenship education, in-service teacher training, teachers’ ability
Promoting Teachers’ Ability in Classroom Discussions About Controversial Issues One important goal of education is to develop critical, democratic citizens who participate actively and responsibly within society (Knight Abowitz, & Harnish, 2006; Bartels, Onstenk, & Veugelers, 2016; Osler & Starkey, 2006; Schuitema, Radstake, Van der Pol, & Veugelers, 2018). Classroom discussions are often regarded as a pivotal means towards reaching this goal (e.g., Knight Abowitz, & Harnish, 2006; Bartels et al., 2016; Haste, 2004; Osler & Starkey, 2006; Schuitema, Ten Dam, & Veugelers, 2008; Schuitema et al., 2018), through facilitation of the development of attitudes and skills that are characteristic of a responsible and critical democratic citizen (Berkowitz, Althof, Turner, & Bloch, 2008; Geboers, Geijsel, Admiraal, & Ten Dam, 2013; Schuitema et al., 2008; Solomon, Watson, & Battistich, 2001).
However, in practice, teachers often refrain from having classroom discussions about controversial issues (e.g., Bartels et al., 2016; Galston, 2004; Molinari, Mameli, & Gnisci, 2013; Oulton, Day, Dilion, & Grace, 2004; Nystrand, Wu, Gamoran, Zeiser, & Long, 2003; Wilson, Haas, Laughlin, & Sulan, 2002). This is possibly because many teachers feel distrusted or unsure about their ability to engage students in such activities, as is found in most studies (e.g., Byford, Lennon, & Russel, 2009; Oulton et al., 2004; Radstake & Leeman, 2010). Yet, to our knowledge, few studies have investigated how teachers’ ability in teaching controversial issues could be promoted (Nucci, Creane, & Powers, 2015). Therefore, in the present study, we sought to investigate if and how an in-service teacher training program could enhance teachers’ ability in classroom discussions about controversial issues that promote democratic citizenship.
Democratic Citizenship Education
Citizenship has become an obligatory school subject in many countries, including almost every European country (Eurydice, 2005), the US (Torney-Purta, Lehmann, Oswald,
& Schultz, 2001), Australia (Print & Gray, 2000), and Canada (Osborne, 2001). Reasons for compulsory citizenship education worldwide are an increase in multiculturalism, growing individualism, and a decrease in active civic participation, especially among younger generations (Nelson & Kerr, 2006). For similar reasons, in the Netherlands, schools are obligated to integrate citizenship education into their curriculum since 2006 (Bron & Thijs, 2011).
Visions on citizenship can vary substantially depending on political views and/or the characteristics of the particular society (Ten Dam, Geijsel, Reumerman, & Ledoux, 2011). Many authors from democratic societies argue that citizenship education should focus on providing students with knowledge, skills, and attitudes that are necessary to function well within a democratic society (e.g., Glass, 2000; Haste, 2004; Holmes, 2001; Kerr, 1999; Naval, Print, & Veldhuis, 2002; Print & Coleman, 2003; Torney-Purta, 2004). Furthermore, it is argued that democratic citizens should be able to make critical contributions to society (e.g., Lipman, 1991; Ten Dam & Volman, 2004; Wardekker, 2001; Weinstein, 1991; Westheimer, 2008). This means that democratic citizens can critically evaluate varying perspectives and are able to critically reflect on the justification of current policies and possible alternatives based on democratic values (Westheimer, 2008). In addition, it is suggested that democratic citizens should be active citizens, meaning that they participate in dialogue with other citizens, and engage in public debate and actions (Bartels et al., 2016).
Knowledge that is assumed to be necessary to become a democratic citizen entails knowledge about the functioning of a democratic society, such as knowledge of the constitution, civil rights, and government (Hicks, 2001; Kerr, 1999). Considering skills, argumentation and communication skills are seen as important (Bartels et al., 2016;
Battistoni, 1997; Schuitema, Van Boxtel, Veugelers, & Ten Dam, 2011). Last, attitudes that are considered to be valuable for a democratic society are tolerance, respect, responsibility,
appreciation of people’s differences, and social involvement (Grant, 1996; Cogan & Morris, 2001).
Classroom Discussions and Democratic Citizenship
It is often assumed that schools can contribute to the development of democratic citizens through classroom discussions about controversial issues (e.g., Knight Abowitz, & Harnish, 2006; Bartels et al., 2016; Haste, 2004; Osler & Starkey, 2006; Schuitema et al., 2008, 2018; Waterson, 2009). Namely, many researchers have suggested that classroom discussions about controversial issues could facilitate the development of attitudes and skills that are required for democratic citizenship (Berkowitz et al., 2008; Geboers et al., 2013; Schuitema et al., 2008; Solomon et al., 2001). More specifically, it is argued that classroom discussions stimulate students’ critical thinking and reasoning (e.g., Chinn, O’Donell, & Jinks, 2000; Ten Dam & Volman, 2004; Veugelers, 2000) by providing students with
knowledge about different perspectives and by stimulating students to form their own opinion on social and moral issues (Schuitema et al., 2009). In addition, Parker and Hess (2001) argue, classroom discussions also serve the purpose of teaching students how to participate in a discussion. This is essential as discussions play an important role in democratic societies (e.g., Althof & Berkowitz, 2006; Gutmann & Thompson, 2004; Haste, 2004; Knight Abowitz & Harnish, 2006; Osler & Starkey, 2006; Parker, 2010; Schuitema et al., 2008; Veugelers, 2007). Concerning this purpose, classroom discussions could help students to develop skills and attitudes needed to formulate their own viewpoints, listen to others, and to react critically though respectfully to their peers.
Several large-scale studies revealed that classrooms perceived as open to discussions about controversial issues were positively related to students’ civic knowledge and
participation in political debate (Godfrey & Grayman, 2014; Manganelli, Lucidi, & Alivernini, 2015; Schulz, Ainley, Fraillon, Kerr, & Losito, 2010; Torney-Purta, 2002;
Torney-Purta et al., 2001). Another retrospective study (Keeter, Zukin, Andolina, & Jenkins, 2002) among American youth (age: 15-25, N = 3,246) indicated that classroom discussions about controversial issues were related to students’ active citizenship after school. Keeter et al. (2002) found that students who participated in classroom discussions at high school participated in demonstrations more often, followed political news more frequently, and signed more petitions after high school compared to students who did not participate in classroom discussions about controversial issues. Furthermore, multiple studies revealed that classroom discussions are positively related to students’ critical thinking and reasoning (Berkowitz et al., 2008; Blatt & Kohlberg, 1975; Nucci et al., 2015). Yet, many researchers (e.g., Alexander, 2008; Avery, Levy, & Simmons, 2013; Burbules, 1993; Hess & Avery, 2008; Schuitema et al., 2011; Ten Dam & Volman, 2004; Veugelers, 2000) have argued that classroom discussions about controversial issues should meet certain criteria in order to contribute to democratic citizenship goals. Three important criteria the literature points to will be discussed in the following paragraphs: an open and safe classroom climate,
co-construction, and critical thinking. Open and Safe Classroom Climate
Open classroom climate. First, it is important that classroom discussions are open (Hess & Avery, 2008; Veugelers, 2000), that is, all students should be able to contribute to the discussion as equal participants. This means that the discussion should be open to all
perspectives; all students must be able to freely present their viewpoints, and these viewpoints should be considered by the other students.
Others have argued that it might not be enough that students feel free to bring in their view, but that students should actively participate in the discussion, so that they can improve their discussion- and reasoning skills (e.g., Alexander, 2008; Brown & Renshaw, 2000; Chinn et al., 2000). As such, it is argued that teachers should stimulate equal participation and
prevent the discussion from being dominated by a small group of students (Burbules, 1993; Kumpulainen & Kaartinen, 2003). One way for the teacher to enhance equal participation may be to actively point out students to share their point of view, instead of allowing students to share their views voluntarily (Dallimore, Hertensein, & Platt, 2004).
Moreover, it is important that classroom discussions about controversial issues are not dominated by the teacher himself/herself. Whereas teachers are commonly regarded as experts on the subject matter (Hand & Levinson, 2012) who regulate the content and process of classroom interaction (Molinari et al., 2013), classroom discussions about controversial issues ask for a teacher to be less directive than in common educational practice, and to make more room for students’ contributions.
Safe classroom climate. When discussing social issues in the classroom, it is likely that students feel personally or emotionally attached to these issues (Hess & Avery, 2008). In addition, it is also likely that students will share personal experiences (Russel et al., 2008). As such, it is important that, like in all other educational subjects, teachers create a safe
environment, in which all members feel heard and respected. To realize such an environment, it is important that the teacher stimulates and praises respect between all participants of the discussion (Fischer, 2000).
The safety and openness of the classroom climate cannot be fully distinguished. Rather, when a classroom climate is not safe, it is probably also not open. Namely, when the classroom environment is not respectful towards all perspectives, some students will not feel safe to share their opinion (Russel et al., 2008). As a consequence, full active participation will not be satisfied and the value of the classroom discussion will be lost for students who do not participate in the discussion. Similarly, the value of the classroom discussion will also be diminished for students who do participate, as relevant perspectives are left out the
manner with others who hold different views, which is also a goal of classroom discussions (Hess, 2009; Parker & Hess, 2001; Scott, Mortimer, & Aguilar, 2006).
Literature points out some concrete ways for teachers to stimulate a safe and respectful environment for a classroom discussion. First of all, many researchers have suggested the use of ground rules for discussion (e.g., Hancock, 2018; Russel et al., 2008; Schuitema et al., 2011). Russel et al. (2008) provide some examples, such as ‘every student has the right to contribute to the discussion without being interrupted,’ ‘contributions should be free from hurtful language,’ and ‘non-verbal communication should also be respectful (e.g., no eye-rolling).’
However, ground rules do not guarantee respectful communication between students. Whenever students do use hurtful language or prejudice, it is if important that the teacher intervenes and explicitly discourages the inappropriate contributions (Hancock, 2018; Russel et al., 2008), although sometimes it may be hard for teachers to draw a line between
contributions that are appropriate and contributions that are not. Co-Construction
It is not only important that students feel safe to share their view during classroom discussion about controversial issues, but also that the students listen and respond to each other (Alexander, 2008; Burbules, 1993). Without responding, there is no room for co-construction. Co-construction occurs when individuals reflect and elaborate on each other’s input and is assumed to enhance people’s ability to reason (Berkowitz et al., 2008;
Kumpulainen & Kaartinen, 2003; Rojas-Drummond & Mercer, 2003; Van Boxtel, 2004; Webb, 2009). Berkowitz and Gibbs (1983) found evidence for the importance of co-construction for students’ moral reasoning. An analysis of 30 dialogues revealed that the occurrence of statements that reflected or elaborated on the argumentation of their fellow student, enhanced students’ moral reasoning. This study implies that it is important that
teachers stimulate students to respond to each other or ask each other questions. This may be especially important when the discussion shows little depth or when students deviate from the topic or try to dodge questions or criticisms from fellow students.
Critical Thinking
It is believed that critical thinking is essential for citizens to participate in a democratic society, and to make their own contribution to the society (Miedema & Wardekker, 1999; Ten Dam & Volman, 2003). As such, critical thinking has become one of the aims of democratic citizenship (Ten Dam & Volman, 2004). In general terms, critical thinking is defined as ‘reasonable reflective thinking that is focused on deciding what to believe or do’ (Ennis, 1991, pp.1-2). According to Ennis, critical thinking requires both skills and attitudes. Skills that are needed are related to argumentation (e.g., judging credibility of sources, analyzing arguments) and attitudes are related to open-mindedness (e.g., willingness to take into account alternative perspectives and to be well informed). As such, in order to promote critical
thinking it is argued that classroom discussions require a focus on arguments (e.g., Schuitema et al., 2011), and different perspectives (e.g., Avery et al., 2013).
Thus, to promote critical thinking in the context of democratic citizenship education, first it is important that students are stimulated to reason and argument on complex, societal, or controversial issues. Chinn et al. (2000) state that the quality of argumentation is dependent on the argumentation complexity. Simple argumentation, they argue, refers to standpoints supported by isolated reasons. Argumentation becomes more complex when positions are supported by multiple arguments or evidence, or complemented with rebutted
counterarguments. Argumentation becomes even more complex when given arguments are supported with additional arguments, or integrated with each other. As such, to stimulate critical thinking through classroom discussions, teachers could encourage students to formulate arguments, reflect on given arguments, and seek additional (counter)arguments.
In addition, to think critically, students need to integrate multiple perspectives into their argumentation. Therefore, for a high quality classroom discussion, it is essential that varying perspectives are brought in and explored (Avery et al., 2013; Banks, 2004; Schuitema et al., 2009, 2011). By becoming acquainted with different perspectives, students will develop richer, more comprehensive, and more accurate understandings of the issues at hand, which will be reflected in more nuanced opinions and more efficient communication with others who hold different views (Avery et al., 2013; Gasiorek & Hubbard, 2017, Tjosvold & Johnson, 1977). Indeed, a study by Schuitema et al. (2018) suggested that teachers could enhance the quality of the discussion by introducing more perspectives. From this, it follows that teachers need to have a good understanding of the issue at hand, so that they are able to recognize which perspectives are being discussed and which perspectives can still be
introduced (Hess & Avery, 2008; Parker, 2003; Radstake & Leeman, 2010; Schuitema et al., 2009, 2011).
Furthermore, when discussing societal issues it is important that students think critically about what one should think or how one should act in a particular situation
(Rombout, Schuitema, & Volman, 2004; Zeidler & Sadler, 2007). As such, students need to make moral judgments about what is right and what is wrong (Zeidler & Sadler, 2007). Therefore, within citizenship education students should reflect on moral values and integrate them in their personal positions (Weyringer, Patry, & Weinberger, 2012). Moral values can be defined as conceptual ideas about how people should live together (Rokeach, 1973;
Veugelers, 2007) and can be distinguished from social norms or laws in that moral values do not apply to specific situations but have a much more abstract nature (Killen, 2007). One study by Schuitema et al. (2011) among pre-university students (age 13 to 14) showed that students who referred to moral values during discussions showed higher quality in their
argumentation in a written essay. As such, teachers could stimulate students to integrate moral values in their reasoning during classroom discussions.
Teacher Roles in Classroom Discussions about Controversial Issues
Combined, the criteria mentioned above make classroom discussions about
controversial issues a very demanding task for teachers (Parker & Hess, 2001; Radstake & Leeman, 2010). Especially because the three criteria discussed above may require teachers to take on different roles and to balance between those roles. For example, a neutral, facilitating teacher role is expected to make students perceive the classroom climate as open (Russel, Soysa, Wagoner, & Dawson, 2008). Yet, on the other hand, in some cases the teacher may need to step away from a facilitating role. For instance, to intervene students who make inappropriate contributions that jeopardize the safety of the discussion (Russel et al., 2008). Thus, in these situations, the teacher has to balance between guarantying that all participants feel safe and respected, and enabling a discussion that is open to all perspectives.
Moreover, the teacher may need to step away from a facilitating role to stimulate critical thinking. The teacher could do this by making students reflect on arguments of themselves and others, or by introducing new perspectives (Hess & Avery, 2008; Parker, 2003; Radstake & Leeman, 2010; Schuitema et al., 2009, 2011). Last, teachers may choose to step away from a facilitating role to stimulate reflection on (moral or democratic) values (Veugelers, 2000). Yet, when the teacher stimulates or provides critique on given arguments or discussed values, students may obtain the impression that the teacher is in favor of a particular view that is in conflict with their own view, and consequently feel less comfortable to share their personal viewpoint. In line with this, Russel et al. (2008) suggested that teachers should not share their own personal position on the controversial issue at hand, at least not at the start of the
discussion. Yet, on the other hand, it might be relevant for teachers to share their opinion, to model good argumentation, to introduce a new perspective, or to promote engagement of the
students. Still, it remains important that all students feel free to bring in their view even when this view differs from the view of (most of the) other students and/or the teacher. As such, teachers need to balance between different teacher roles to create a discussion that is both critical as well as open and safe.
Teachers’ Ability in Classroom Discussions about Controversial Issues
Using online surveys among 1,117 teachers from Dutch primary and secondary schools, Sijbers, Elfering, Lubbers, Scheepers, and Wolbers (2015) found that most teachers experienced no difficulties in discussing controversial issues. Yet, many studies revealed that teachers often refrain from teaching controversial issues (e.g., Bartels et al., 2016; Galston, 2004; Molinari et al., 2013; Oulton et al., 2004; Nystrand et al., 2003; Wilson et al., 2002). Moreover, in contrast to the study by Sijbers et al. (2015), most studies indicate that many teachers do feel unsure or distrustful of their ability to teach controversial issues (e.g., Byford et al., 2009; Oulton et al., 2004; Radstake & Leeman, 2010). Oulton et al. (2004) found that most of the teachers in English primary and secondary schools had no formal training about teaching controversial issues, and the majority of the teachers reported that they did not feel sufficiently equipped to engage students in classroom discussions about controversial issues. Similarly, a small-scale study on ethnically mixed classrooms in the Netherlands (Radstake & Leeman, 2010) indicated that secondary school teachers did not feel prepared to involve their students in discussions about social justice or (in)equality-related issues. In addition, a study among high school social studies teachers from the United States (Byford et al., 2009) showed that many teachers felt unsure of their ability to facilitate discussions about controversial issues. Reasons for this uncertainty were teachers’ lack of confidence in the curriculum and materials, and concerns about student behaviors that may lead to conflict or disruptions. Furthermore, multiple studies (e.g., Byford et al., 2009; Oulton et al., 2004; Radstake & Leeman, 2010; Sijbers et al., 2015) pointed out that many teachers are unsure what is
expected of them during classroom discussions about controversial issues. As such, in the present study, we investigated how teachers’ ability in facilitating classroom discussions about controversial issues could be promoted.
Promoting Teachers’ Ability in Classroom Discussions about Controversial Issues To our knowledge, few studies have tested an in-service teacher training aiming to enhance teachers’ ability in teaching controversial issues (Nucci et al., 2015). More generally, research on professional development for teachers, has revealed that teachers trainings are most effective when they contain a combination of tools for learning (i.e., strategies, theory, example lessons) and reflection (1) on their own teaching practices (Avalos, 2011; Stewart, 2014) based on a clear conceptual framework (2) (Van de Pol, Brindley, & Higham, 2017). In addition, teacher trainings are found to be most effective when teachers participate in groups (3) (Avalos). Furthermore, it is found to be important that teachers are able to match the training to their own needs (Bayar, 2014), by having a choice in what to learn (4) and how to learn it (Stewart, 2014).
Therefore, in the present study, teachers participated in small groups (3) in a teacher training that provided teachers tools (i.e., knowledge, strategies, and example lessons) for facilitating classroom discussions about controversial issues, as well as opportunities to reflect (1) on lessons in which they practiced with the provided tools. The framework (2) that was used for the tools and reflective activities consisted of the three important criteria of classroom discussions (open and safe discussion climate, co-construction, and critical thinking) and four teacher roles (i.e., maintaining a safe classroom climate, facilitating the discussion, stimulating critical thinking, stimulating moral values). Additionally, teachers set and adapted their personal goals (4), so that the training was relevant for all participating teachers.
Present Study
In the present study, we sought to investigate how teachers’ ability in classroom discussions about controversial issues could be promoted. First, using questionnaires we examined whether an in-service teacher training promoted teachers’ ability in facilitating classroom discussions about controversial issues that meet important criteria (i.e., open and safe classroom climate, co-construction, critical thinking) (1st Research Question). Second, using interviews with teachers who participated in the teacher training, we investigated if and how the training improved their practices in classroom discussions about controversial issues (2nd Research Question). The interviews were expected to provide useful insights into which aspects of the training teachers perceived to be helpful in enhancing their ability to facilitate discussions about controversial issues. In addition, these interviews could point out whether teachers still experienced difficulties during classroom discussions despite the teacher training. These insights could be used to attune future teacher programs better to the difficulties and challenges teachers perceive in teaching controversial issues.
Method Research Design and Sampling Procedure
For the present study, we used a nonrandomized control group pretest-posttest design to test the effectiveness of an in-service teacher training program. First, teachers from
secondary (vocational) education across the Netherlands were approached for participation in the teacher training via personal networks or LinkedIn. Next, teachers who volunteered in the teacher training were asked to approach one of their colleagues from the same school (who taught the same subject to students from the same age and educational level) to participate in the study as a teacher for the control group. All teachers (i.e., training group and control group) filled out a questionnaire before and after the teacher training. In addition, interviews
were conducted with six randomly selected teachers from the training group to investigate how the training affected their ability in teaching controversial issues.
Participants
The final sample consisted of 89 teachers (62.9% female). From the total sample, approximately half of the teachers (n = 46) participated in the teacher training. Teachers on average were 38.9 years old (range = 22 – 64), and had 11 years (range = 1 - 41) of teaching experience. The teachers identified themselves as Dutch (n = 81), Moroccan (n = 3), Turkish, Surinamese, Antillean, Indian, or Polish (n = 1). Subjects that teachers taught were: the Dutch language (n =19), social studies (n = 17), history (n = 15), citizenship (n = 8), the English language (n = 8), geography (n = 6), biology (n = 2), or a combination of these and other subjects. In addition, teachers taught students from different educational streams:
pre-vocational education (n = 36), general secondary education (n = 49), pre-university education (n = 41), and/or secondary vocational education (n = 33). The mean age of the students (54.8% girls) they taught was 15.7 years (range = 11 – 22). Students identified themselves as Dutch (71.3%), Dutch/Turkish (6.4%), Dutch/Moroccan (4.6%), Dutch/Surinam (3.9%), Surinamese (1.5%), or differently (< 1.5% per ethnic identity).
Procedure
Development of the teacher training. The teacher training was designed by experienced teachers, teacher educators, and researchers during a one-year interactive
process. First, a group of teachers with substantial experience in teaching about controversial issues (including both secondary school teachers and university teachers), developed tools (i.e., strategies, and example lessons) which teachers could use during classroom discussions about controversial issues. The group based these tools on personal experiences, scientific literature, and input from an expert panel of school leaders, teacher educators, teachers, staff members of civil organizations, and researchers. Next, in a pilot study, a group of 10 teachers
from different secondary schools experimented with the developed strategies in their classrooms. Based on observations, video-recordings, and interviews from the pilot study, indications of effectiveness and suggestions for improvement were fed back to the
development team, who again used this feedback to improve the tools for the teacher training. Data collection. All teachers (including control group teachers) filled out a
questionnaire (see Instruments) before (September 2018) and after the teacher training (March/April 2019) about their perceived ability in classroom discussions in controversial issues. Teachers on average spent 20-30 minutes filling out the questionnaire.
In addition, after completion of the training, six randomly selected teachers from the training group were asked to facilitate two additional classroom discussions about
controversial issues. These lessons were used to obtain a more detailed understanding of how the training improved teachers’ practices in teaching controversial issues. To facilitate
comparisons, teachers were asked to engage their students in classroom discussions about the next two themes: 1. Immigration, and 2. Freedom of religion. These themes were chosen as these fitted within all subjects of participating teachers. For each theme teachers were
provided with a statement to discuss with their students: 1. Country boarders should not exist; everyone should be able to live where he/she wants, and 2. Police officers should not be allowed to wear a headscarf, as they should be neutral. Shortly after the second discussion, stimulated recall interviews (see Instruments) were conducted with the six teachers from the training group. The interviews on average took 45 minutes and were all recorded (in
agreement with the teachers) for analysis purposes. Teacher Training
Teachers in the training group participated in small groups (3-5 teachers) in three 2-hour meetings led by teacher educators. The teacher training could be characterized as a blended-learning approach. First, the teachers were provided with relevant tools (i.e.,
knowledge, strategies, and example lessons) for teaching controversial issues through a digital learning space. These tools were organized using a clear conceptual framework including three important criteria of classroom discussions (open and safe discussion climate, co-construction, and critical thinking), and four teacher roles (i.e., maintaining a safe classroom climate, facilitating the discussion, stimulating critical thinking, stimulating moral values). Teachers were stimulated to introduce the three criteria to their students when teaching about controversial issues, and to use these criteria to evaluate their lessons. Information about different teacher roles in classroom discussions was provided to show teachers that different roles may be more appropriate in different situations depending on the goal(s) of the
particular discussion. Moreover, different teacher roles were discussed during the meetings to stimulate teachers to reflect on their own role, and possible tensions between the different teacher roles.
Second, using the same conceptual framework described above, during the meetings teachers reflected on their experiences in teaching controversial issues. To optimize the usefulness of the training meetings, teachers were asked to facilitate at least two classroom discussions about controversial issues in between the training meetings, in which they practiced with provided strategies, tools, and materials. Next, during the meetings, the teachers discussed their (joint) experiences from these discussions in order to develop new plans for improving their teaching practices in following classroom discussions.
Additionally, during the meetings and with the assistance of teacher educators, teachers set personal goals for the next two discussions. These goals aimed to make the teacher training relevant for all teachers by adjusting the goals to their personal needs. In addition, these goals aimed to provide teachers a specific focus on what to learn. Based on personal goals, between the meetings, teachers invested time in preparing their own discussions. For example, one teacher decided he wanted to become more capable of
introducing relevant perspectives to the discussion (i.e., personal goal) and therefore, he searched for existing arguments related to the statements he wanted to discuss with his students.
Instruments
Questionnaire. A new questionnaire was designed for the present study. The questionnaire was introduced in a pilot version to a group of biology and social studies teachers in training (N = 56). Based on their suggestions, formulation of some questions was improved. The final questionnaire consisted of 25 statements about teachers’ ability in teaching controversial issues. These 25 statements measured teachers’ perceived ability in; creating an open and safe discussion climate (14 items), stimulating co-construction (five items); and, promoting critical thinking (six items). All statements started with ‘When teaching controversial issues, I am able to’. Some example items are ‘When teaching controversial issues, I am able to engage students in the discussion’ (i.e., open and safe discussion climate), ‘When teaching controversial issues, I am able to let students share ideas with each other’ (co-construction), and ‘When teaching controversial issues, I am able to let students reflect critically on the relevance of their arguments’ (i.e., critical thinking). Teachers responded to the statements on a 10-point scale ranging from 1 = not at all to 10 = very well. Cronbach’s alpha was high (> .87) for all three constructs at both the pre- and post test, as is shown in Table 1.
Table 1
Cronbach’s Alpha Values for Dependent Variables at Pre and Post-Test
Open and Safe Co-Construction Critical Thinking
Pre .93 .87 .86
Background characteristics. The questionnaire also included some questions about teachers’ background characteristics (e.g., gender, age, years of experience, ethnicity, the subjects they taught, and the educational level of their students).
Interviews. Using interviews with teachers who participated in the teacher training (n = 6), we investigated if and how the training improved teachers’ practices in classroom discussions about controversial issues (2nd Research Question). In order to obtain concrete answers to this research question, teachers were asked about their teaching practices in two classroom discussions about statements provided by the research group shortly after
completion of the teacher training. Teachers were asked to elaborate and explain their choices and actions during these classroom discussions, and were asked how these choices and actions were related to the teacher training. As such, the interviewer observed both discussions and noted relevant teaching practices (i.e., practices related to the three criteria, or four teacher roles) and used these as input for the interview. For example, concerning the third criterion (i.e., critical thinking), the interviewer asked some of the teachers: “During the two lessons I noticed that you introduced new perspectives to the discussion frequently, [interviewer names a few examples], is this something you always do, or is this something you learned from the teacher training?” Similarly, the interviewer selected relevant video-fragments from the two classroom discussions and asked teachers to explain why they acted in a particular way, what they thought was the effect of the action on their students, and how this action related to the teacher training. Furthermore, teachers were asked about which role they took during the classroom discussions, whether they perceived any tensions between the different teaching roles, and whether they adapted their role in classroom discussions throughout the teacher training. Finally, teachers were asked whether they still experienced difficulties and/or challenges despite the teacher training, and whether they had any suggestions for improvement of the teacher training.
Analyses
Questionnaires. To examine whether the teacher training enhanced teachers’ ability in classroom discussions about controversial issues (1st Research Question), the teacher-reported questionnaires were analyzed using multilevel linear mixed models for repeated measures. Multilevel models were used as this type of models is suited to deal with nested data (Hox, Moerbeek & Schoot, 2018). In this case, the data was nested as observations were nested within participants. In contrast to many statistical tests, multilevel models correct for dependencies between observations (Hox et al., 2018). Another advantage of multilevel analysis of longitudinal data is that this method uses all available data for each time point and therefore tends to have larger power than repeated measures ANOVA (Hox et al., 2018).
Linear mixed models allow examining linear relationships between both within and between subject variables and a dependent variable. In this case, we examined the effects of a within subject variable time (i.e., pre/post) and a between subject variable condition (i.e., training/control) on the three dependent variables separately (i.e., teachers’ ability in stimulating; an open and safe discussion climate; co-construction; and critical thinking). Moreover, we investigated the interaction effect between time and condition, as we expected to find an effect of time for the training group, but not for the control group.
The significance of the interaction effects was tested by comparing the model fit of the models with and without the interaction effect, using deviance tests (Hox et al., 2018). In addition, for all fixed effects separately, one-sided t-tests were conducted. Whenever a significant interaction between time and condition was found, additional two-sided t-tests were performed to examine the effects of time on the dependent variable for each condition separately, using a Sidak adjustment for multiple testing. For all other tests, a Nominal Type I error rate of α = .05 was used. Furthermore, as measure of effect size, standardized regression
coefficients or Cohen’s d (for follow-up tests) were obtained and interpreted using guidelines of Cohen (1992): negligible < 0.1 < small < 0.3 < moderate < 0.5 < large.
All analyses were conducted in R (version 3.4.2; R Core Team, 2017) using the ‘multilevel’ and ‘nlme’ packages. All models were estimated using unstructured covariance matrices, meaning that the variances of the observations were allowed to differ across the two time points. In addition, all models were estimated using restricted maximum likelihood (REML)1, as this procedure is suited for multilevel models (Corbeil & Searle, 1976). Furthermore, another advantage of REML is that the parameters are estimated using all available data for each time point. As such, REML was useful for the present study, which contained a substantial amount of missing data; 35 teachers did not fill out the questionnaire at the post-test (all teachers filled out the questionnaire at the pre-test). From this group, 31 teachers left the research project. Known reasons for leaving the project were: work overload, illness, maternity leave, or the passing of a family member. To test whether the missingness could be explained by other variables in the dataset, we performed a Little’s (1988) MCAR test in SPSS, including teachers’ scores on the dependent variables (open and safe discussion climate, co-construction, critical thinking) as well as several teacher characteristics (e.g., training condition, age, gender, ethnic identity, years of experience, educational level of their students). This test showed an insignificant result, Δχ2(16) = 4.87, p = .996, providing no evidence that missing data was systematic. Missing data was not imputed as (RE)ML produces more accurate parameter estimates than imputation methods when sum scores are missing entirely (Noorae, Molenberghs, Ormel, & Van den Heuvel, 2018), as was the case in the present study. Namely, teachers responded to either all or none of the statements from the questionnaire at the post-measurement.
1 As REML does not enable comparisons in model fit between two different models, deviance tests were conducted with models that were estimated using unrestricted ML.
Interviews. To examine if and how the training improved teachers’ practices in classroom discussions about controversial issues (2nd Research Question), stimulated recall interviews were conducted with six teachers from the training group. Interviews were coded using the software program ATLAS.ti. Prior to coding, four code groups were determined: training effects (i.e., improvements that the participants made in their teaching practices due to the training program), teacher practices (teacher practices that were not a result of the teacher training), challenges/difficulties, and teacher training suggestions. Next, for every quote that represented a training effect, a teacher practice, a challenge, or a suggestion for the teacher training, it was analyzed whether this quote related to one of the three important criteria of classroom discussions (i.e., open and safe classroom climate, co-construction, critical thinking), teacher roles, discussion methods (added during coding), or something else. This way, the interviews could be used to gain insights into what teaching strategies were perceived as useful, and which challenges may require additional strategies.
Results Questionnaires
First, using questionnaires we examined whether an in-service teacher training promoted teachers’ ability in classroom discussions about controversial issues (1st Research Question).
Descriptive Statistics. Table 2 shows the correlations among teachers’ ability scores on all three dependent variables (i.e., Open and Safe, Co-Construction, and Critical Thinking) for both the pre- and post measurement. This table reveals that the three variables correlated strongly and positively with each other at the pre-measurement (rs ranged from .76 to .82). Furthermore, teachers’ ability rates within each variable correlated moderately to highly and positively among the two time points (rs ranged from .46 to .57).
Table 2
Correlations Among Teachers’ Ability Scores
1 2 3 4 5 6
Open and Safe
1. Pre - 2. Post .57** - Co-Construction 3. Pre .82** .30* - 4. Post .54** .67** .48** - Critical Thinking 5. Pre .79** .44** .76** .50** - 6. Post .50** .76** .34* .70** .57** - * p < .05 ** p < .01
Table 3 shows all means of teachers’ ability scores for teachers from the control and training condition separately. Teachers’ scores on the pre-test ranged from 6.42 to 7.32 out a maximum of 10. Next, to see whether the teacher training was effective, multilevel linear mixed models were estimated. These results will be discussed for Open and Safe Classroom Climate, Co-Construction, and Critical Thinking, respectively.
Table 3
Means, Standard Deviations, and Group Sizes of Teachers’ Ability Scores
Critical Thinking
Pre Post
Condition M SD N M SD N Control 7.05 0.93 43 7.26 1.03 27 Training 6.90 1.12 46 7.35 0.62 27
Open and Safe Co-Construction
Pre Post Pre Post
Condition M SD N M SD N M SD N M SD N Control 7.32 0.97 43 7.48 1.03 27 6.97 1.01 43 7.27 0.94 27 Training 6.91 1.09 46 7.71 0.62 27 6.42 1.33 46 7.33 0.79 27
Open and Safe Classroom Climate. For Open and Safe Classroom Climate, we first estimated a multilevel intercepts-only model to calculate the intra class correlation (ICC). The ICC turned out to be .61, indicating the average correlation between two observations from the same participant (Hox et al., 2018). This substantial amount of dependency in the data confirmed the appropriateness of a multilevel model.
As such, the hypothesized model was estimated including one fixed intercept, two conditional effects of time and condition (i.e., training/control group), an interaction between time and condition, and a random intercept. This model explained 6.21% more variance in Open and Safe Discussion Climate than the intercepts-only model, which turned out to be a significant increase, Δχ2(3) = 21.01, p < .001. The results of the hypothesize model are displayed in Table 4.
Table 4
Results of Fixed and Random Effects of Model for Open and Safe Discussion Climate
*Standard errors of random effects were not provided by R as the lower bound of the variance estimate(s) was (were) below zero.
Model 1
Fixed effects Est. s.e. p
Intercept 7.32 0.15 < .001
Time 0.07 0.15 .635
Condition -0.41 0.21 .056
Time*Condition 0.62 0.21 .005
Random Effects Est. s.e.
Level-1 variance 0.32 *
Level-2 variance 0.67 *
Deviance 375.13
To test whether the interaction effect between time and condition was significant, we compared the fit of this model to a similar model without the interaction effect. This
comparison indicated that the model including the interaction effect fitted the data
significantly better than the model without the interaction effect, Δχ2(1) = 8.37, p = .004. The interaction effect turned out to be large and positive, β = .61. The regression coefficient, b = 0.62, could best be interpreted using a plot of the data (see Figure 1).
Figure 1 showed that the training group seemed to improve towards the post-test in contrast to the control group that seemed to show no change. Follow-up comparisons (Table 5) confirmed this conclusion. The training group increased from 6.91 at the pre-test to 7.60 at the post-test, which indicated a significant mean difference, ΔM = 0.69, t(47) = -4.65, p < .001, whereas the control group remained stable over time, ΔM = 0.07, t(47) = -0.48, p =.635. The effect of time for the training group appeared to be mediocre, Cohen’s d = .39. Finally, the effect of Condition indicated that teachers from the training group did not score
significantly lower on Open and Safe Classroom Climate at the pre-test compared to the control group, t(87) = 1.94, p = .056, although the p-value was close to the criterion of .05.
Figure 1. Teachers’ ability in stimulating an Open and Safe Classroom Climate for pre- and post measurement.
Table 5
Results of Follow-Up Comparisons for Open and Safe Discussion Climate
Co-Construction. For the second dependent variable, Co-Construction, again first a multilevel intercepts-only model was estimated to calculate the ICC. Similar to the first dependent variable, substantial dependency was found in the data, ICC = .50. Next, the hypothesized model was estimated including the same parameters as for the first dependent variable Open and Safe. This model (Table 6) explained 8.89% more variance in Co-Construction compared to the intercepts-only model, Δχ2(3) = 19.97, p < .001.
Table 6
Results of Fixed and Random Effects of Model for Co-Construction
Model 2
Fixed effects Est. s.e. p
Intercept 6.97 0.17 < .001
Time 0.19 0.19 .315
Condition -0.55 0.23 .019
Time*Condition 0.59 0.27 .032
Random Effects Est. s.e.
Level-1 variance 0.54 0.26
Level-2 variance 0.65 0.30
Deviance 414.64
ICC .55
Condition M Pre (SE) M Post (SE) ΔM t(52) p Cohen’s d Control 7.32 (0.15) 7.39 (0.17) 0.07 -0.48 .635 .04
A model fit comparison between the hypothesized model and a similar model without an interaction between time and condition, revealed that the model including an interaction effect fitted the data significantly better, Δχ2(1) = 4.82, p = .028. Additionally, the interaction effect appeared to be mediocre and positive, β = .53. The regression coefficient of the
interaction, b = .59, was again interpreted using an interaction plot (see Figure 2). The interaction plot (Figure 2) showed a similar pattern as for the first dependent variable; teachers in the training group showed an increase, whereas control group teachers appeared to report similar scores across the two time points. This pattern was confirmed by follow-up tests (Table 7); teachers from the training group increased from 6.42 on the pre-test to 7.21 at the post-test, which indicated a significant change, ΔM = 0.79, t(52) = -4.14, p < .001. This effect could be considered to be mediocre, Cohen’s d = 0.35. To the contrary, teachers from the control group showed no significant change in their ability of
Co-Construction, ΔM = 0.19, t(47) = -1.02, p = .315. Finally, the effect of Condition showed that teachers in the training group perceived their ability in stimulating Co-Construction
significantly lower than teachers in the control group at the pre-test, b = -0.55, t(87) = -2.40, p = .019.
Table 7
Results of Follow-Up Comparisons for Co-Construction
Condition M Pre (SE) M Post (SE) ΔM t(52) p Cohen’s d Control 6.97 (0.17) 7.17 (0.20) 0.19 -1.02 .315 .08
Training 6.42 (0.16) 7.21 (0.20) 0.79 -4.14 < .001 .35
Critical Thinking. For the third dependent variable, Critical Thinking, again a significant amount of dependency in the data was found, ICC = .60. The parameter estimates of the hypothesized model for Critical Thinking (Model 3) are summarized in Table 8. This model explained only 1.15% more variance in Critical Thinking than the intercepts-only model, which turned out to be no significant difference, Δχ2(3) = 7.76, p = .051. In addition no significant difference was found in model fit between the hypothesized model and a similar model without an interaction between time and condition, Δχ2(1) = 1.94, p = .163. This means that the effect of time on Critical Thinking did not depend on the condition of the teacher (i.e., control or training). Furthermore, Table 8 shows that neither the effect of time, b = 0.11, t(52) = 0.73, p = .470, nor the effect of condition, b = -0.16, t(87) = -0.74, p = .459, was statistically significant. These results indicate that teachers’ scores on Critical Thinking did not change over time, nor differed between the control and training group.
Table 8
Results of Fixed and Random Effects of Key Models for Critical Thinking
* Standard errors of random effects were not provided by R as the lower bound of the variance estimate(s) was (were) below zero.
Interviews
Second, using stimulated recall interviews with teachers who participated in the teacher training (n = 6), we investigated if and how the teachers improved their practices in classroom discussions about controversial issues due to the training program (2nd Research Question). First, we will provide some background information about the interviewed teachers, followed by information about which aspects of classroom discussions teachers chose to work on during the training program. Next, we will discuss on which aspects of classroom discussions about controversial issues teachers improved, as well as which
challenges/difficulties teachers still experienced despite participating in the teacher training. For all interviewed teachers, background information is provided in Table 9. Table 9 shows that the teachers (66.6% female) were between 28 and 55 years old, and had between four and 25 years of teaching experience. Together teachers taught four different subjects
Model 3
Fixed effects Est. s.e. p
Intercept 7.05 0.15 < .001
Time 0.11 0.16 .470
Condition -0.16 0.459 .441
Time*Condition 0.31 0.22 .171
Random Effects Est. s.e.
Level-1 variance 0.36 *
Level-2 variance 0.62 *
Deviance 378.27
within three different educational streams. Furthermore, Table 9 shows that most teachers taught to classrooms with limited ethnic diversity. In relative terms, Teacher 4 taught to the most diverse group, yet in absolute terms the diversity in his classroom was not so large as all his students were boys, and only three different ethnic groups (e.g., Dutch, Canadian,
Surinam) were represented in his rather small group of six students. As such, Teacher 2 might have taught to the most ethnic diverse group of students, with 27% of the students assigning themselves an ethnic identity other than/additional to Dutch (e.g., Armenian, Surinam, English, Asian, or Kurdish). Notably, the ethnic diversity in the classrooms of the random sample of interviewed teachers was relatively small compared to other teachers who participated in the present study that taught to groups in which up to 88% of the students assigned themselves an ethnic identity other than/additional to Dutch.
When teachers were asked about which aspects of classroom discussions they
primarily worked on during the training, Teacher 1 and 3 said they tried to create more equal participation, and Teacher 2 and 4 indicated they practiced with stimulating critical thinking. As Teacher 5 only just started with having whole group discussions, she worked on all aspects of classroom discussions, and did not really focus on specific aspects of the training. Last, Teacher 6 told that she had not found the time to really practice with any of the tools or materials.
Table 9
Background Information About Interviewed Teachers
Teacher Gender Age Subject Educational Stream (Grade) Years of Teaching Experience Ethnic Diversity*
1 Female 56 Dutch GSE/P-U (1) 16 17%
2 Male 32 Dutch P-U (3) 4 27%
3 Female 28 Citizenship SVE 6 15%
4 Male 42 Citizenship SVE 4 33%
5 Female 28 English P-U (6) 4 19%
6 Female 55 Social Studies P-U (5) 25 19% Note. GSE = General Secondary Education; P-U = Pre-University Education; SVE = Secondary Vocational Education. Ethnic Identity = percentage of students who assigned themselves an ethnic identity other than/additional to Dutch.
Open and safe classroom climate. Most of the teachers perceived the classroom environment as safe during discussions about controversial issues, despite the fact that they did not put much attention and/or effort into stimulating a safe environment. The teachers indicated that their students listened and responded respectfully to each other. They explained that this was because their students could get along well, and/or because their students knew each other for a long time (i.e., multiple years). In contrast, Teacher 1 stated that, as her students were only together very shortly, she had used suggestions from the training materials to start the first discussion in smaller subgroups to create more safety. Yet, Teacher 1 doubted whether this method had been effective and she noticed that the discussion method had a negative side effect; students were sometimes discussing topics that were not related to the issue being discussed.
students, she decided to implement some strategies from the training to make students feel safe to participate in this relatively new form. First, she introduced criteria for evaluation of the discussion to the students: 1. we respond respectfully to each other, 2. we participate and share our view, 3. we listen and respond to each other, 4. we provide arguments and
investigate their credibility, 5. our teacher ensures that different perspectives are brought in to the discussion. Teacher 5 explained that introducing the criteria to the students was helpful as she could remind the students of the criteria at relevant moments, instead of correcting
individual students frequently. Second, for the first discussion Teacher 5 conducted an anonymous poll among her students to investigate whether students were in or out of favor with the particular statement. This way, Teacher 5 explained that, students could see that several classmates shared their view, and consequently, feel safer to share their viewpoint. Last, Teacher 5 stated that she made use of a discussion method that was discussed during the training meetings, which increased both the openness and safety of the discussion climate. For this so-called Fish Bowl method students were placed in an outer circle, or a smaller inner circle. Teacher 5 explained that this discussion method increased the openness of the discussion as all students had an active role: students in the inner circle are having a
discussion and students in the outer circle have the role of active listener, are allowed to ask clarifying questions, and can jump in to the discussion when they want to. In addition, this discussion method makes the students feel safe as they are having a discussion in a smaller group. Teacher 5 stated that the students ‘liked to have a discussion in a smaller group so that they didn’t have the feeling they were making their point to the whole group.’
Although most of the teachers perceived their classroom environment as safe during the discussions they had, several teachers expressed concerns about their ability to maintain a safe classroom climate during discussions about controversial issues in some specific
climate when the discussed topic may be emotionally charged for one or more students. For instance, Teacher 4 mentioned that the training provided him insufficient tools for how to discuss a controversial issue such as homosexuality:
“Those very sensitive issues I talked about before [i.e., homosexuality], I wanted to have more tools for that, as I still find that very difficult. So, how could I discuss some issue properly, when I know this issue does concern the group, without damaging the particular person and without giving that person the feeling that he or she finds him/herself in an unsafe situation?”
Teacher 4 suggested that his uncertainty to discuss homosexuality might have less to do with his ability to stimulate a critical discussion, but rather with a lack of required pedagogical skills to create a discussion that is open and safe for all students. As a consequence, Teacher 4 stated that he rather refrains from discussing homosexuality at all.
Furthermore, some teachers expressed concerns about maintaining a safe classroom climate when a controversial issue might pop up spontaneously (i.e., when no classroom discussion was planned). Teachers stated that they expected or hoped that the training would provide them more tools for how to deal with such situations. Teacher 1 explained that, as she felt uncertain about her ability to deal with such situations, she did not benefit much from the training in which she had to choose the topics and prepare the discussions in advance.
Finally, two teachers expressed concerns about maintain a safe discussion climate when extreme views would turn up. Teacher 5 explained that she would have hoped to get more tools for how to respond to students that ‘have an opinion that doesn’t fit within the norms and values we have here in the Dutch society or within the climate of the school.’ Interestingly, both teachers indicated that unexpected or extreme situations rarely or never occurred within their schools. Teacher 1 suggested that unexpected or extreme situations might occur more often within more diverse classroom compositions.
Concerning stimulating an open discussion climate, four out of six teachers indicated that they put more effort towards letting all students speak up during the discussions
compared to the start of the training. As a consequence, most teachers felt the training made them more capable of stimulating equal participation. For instance, Teacher 3 indicated:
“I think I have improved in distributing turns, approaching all students. I may have been hesitated to do so earlier, but now I do try to engage the whole group in the discussion. I’m more aware of the group who is silent, and I try to involve them in the discussion by asking about their opinions.”
Teacher 5 used a slightly different method to engage all students in the discussion. She provided all students three ‘tickets’ to make a contribution to the discussion and encouraged students to use at least one. Teacher 4 explained that he experienced clear benefits from his increased effort to stimulate students to share their opinion; he obtained more knowledge about the perspectives of his students, and some students who were a bit shy in the beginning appeared to be able to make very useful contributions to the discussions:
“For example, one boy appeared to be very good at putting things into perspective and he even started to ask questions to the other students, and this way he kind of took over my role as a facilitator. Well, isn’t that a very cool outcome of asking ‘What do you think of this all’?”
Some teachers argued that, although stimulating participation in classroom discussions is important, it might not be necessary that all students speak up during a classroom
discussion. For example, Teacher 3 argued that students can also learn and participate by just listening. In addition, Teacher 1 argued that focusing too strongly on making everybody say something may reduce possibilities for co-construction, as time is often limited:
“Now, they all [the students] bring in one thing, as I think everybody should be able to give his opinion. However, this way the time is too limited to let students respond
to each other.”
Although multiple teachers indicated that the training helped them create a more open classroom climate, most teachers still found it difficult to facilitate a discussion that is open to all students, due to the large number of students and/or time limitations. For example, Teacher 1 stated that due to the large group and limited time she could not let all students share their view or make room to discuss their view more deeply. Likewise, Teacher 2 explained that some students who might hold very interesting views get easily pushed into the background in a large group. In addition, Teacher 4 explained that he was better able ‘to give all students the feeling they are a part of the discussion’ in smaller groups. Finally, two teachers mentioned that they found it difficult to keep students engaged with the topic when having a discussion with a large group of students.
Co-construction. None of the teachers reported any improvement in their practices concerning stimulating co-construction during classroom discussions. This is likely because none of the interviewed teachers practiced with stimulating co-construction, but instead focused on other aspects of the training.
Critical thinking. One of the teachers experienced a clear improvement in his practices to stimulate critical thinking. Teacher 4 indicated that the training sessions had helped him to be more persistent when asking students questions, and thereby pushing students to think more critically about why they hold certain views. Teacher 4 explained that persistent questioning makes his students ‘think harder, and it makes them realize that they cannot simply ventilate just about anything.’
Furthermore, Teacher 2 reported that he tried to deepen his knowledge about the topics of the discussions he held with his students to confront students with alternative views or arguments. Yet, Teacher 2 mentioned that he still found it difficult to introduce relevant alternative perspectives to the discussions. He explained that he found it sometimes hard to
consider how different groups of people would think about the controversial issue at stake. Although the other teachers did not learn about the importance of introducing new perspectives to a classroom discussion during the training, for most of the teachers this turned out be something they do regularly when teaching controversial issues. Most of the teachers did not prepare these perspectives prior to the discussions, but introduced these perspectives spontaneously when they felt that this was relevant for the discussion that was going on. In addition, Teacher 3 and 4 stated that they sometimes used videos (e.g., news items, or documentaries) to introduce new perspectives to their students. Most teachers found
introducing alternative perspectives a useful practice to promote critical thinking in students. More specifically, Teacher 3 argued that introducing different perspectives broadened and nuanced the views of her students.
Still, several teachers indicated that it could sometimes be hard to stimulate students to think critically about a controversial issue, especially when few contrasting views are present within the classroom. Teacher 1 told:
“I wanted to place two groups across from each other. But then they all walked towards one side of the classroom, apart from one or two students I believe. Then it becomes a little harder to discuss. Than it’s more like highlighting different opinions, but that’s not really a discussion.”
Similarly, Teacher 3 explained that, although introducing new perspectives to the discussion often helps, it sometimes remains difficult to stimulate a discussion when the group holds similar views:
“I teach for six years now to these groups and from the start I experienced that students will pick the same side. And then that’s it, and the discussion is gone. And then you have to find a solution for that. But still I find it difficult sometimes. To make discussions go on longer. And then it’s over in 10 minutes, and that’s a shame.”
Teacher Roles. Most teachers indicated that the training made them more aware of the existence, benefits, and drawbacks of different roles one can take on as a teacher during a classroom discussion. Also, most teachers indicated that the training triggered them to rethink their own role in classroom discussions about controversial issues. As a result, four teachers indicated that the training had made them take in either a more neutral, or a more steering role depending on the reflection on their role within classroom discussions. As such, Teacher 3 indicated that the teacher training had made her share her own opinion more often with students. She explained that a discussion between one of the other teachers and the teacher educator who led the training sessions had made her think more about whether a teacher should be allowed to share his/her opinion. Teacher 3 concluded that ‘it should be OK to share your opinion as a teacher, as long as you also discuss the other side of the story. Still, Teacher 3 said only to give her opinion when students asked about it, as she thought it was important to be as neutral as possible. She explained, that providing her opinion might hamper the openness of the discussion:
“I still try to be as neutral as possible, as I want to provide students with the possibility to share their view. And sometimes I’m afraid that when I say I totally disagree on something, they [the students] will pick the same side.”
In contrast, Teacher 5 noted that she had placed herself in a more neutral, facilitating role due to the training. She argued that, although this role may not be better than her initial role per se, it did made het more confident in teaching controversial issues. She explained: ‘[in my previous role] I noticed that I became an active participant of the discussion. And students looked at me for input so to speak. And I was always very hesitant about this, is this good or not?’
Similarly, Teacher 4 stated that his role during discussions had become more neutral during the training, as he concluded he might have shared his opinion too much in earlier
discussions. Teacher 4 tried to take in a more neutral role by facilitating the discussion using questions. Teacher 4 argued that this adaptation was positive, as ‘it should not be the role of the teacher to convince his students of his opinion, but to support his students to develop their own opinion’. In addition, and like Teacher 3, he argued that he feared that students would refrain from participating in the discussion when it is clear to them how the teacher feels about the discussed topic. Still, Teacher 4 expressed struggles about how to balance between his new, more neutral, facilitating role and his role to stimulate critical thinking:
“Well, especially with the immigration debate, and the fact that right-wing boys from my class always say ‘those immigrants,’ I find that difficult as, emotionally, I feel like I need to say something about it. But I don’t need to correct them, or well maybe a little,
nurturing is also a part of an educator’s job. But I found that hard, because, well, I do have an opinion on that. … So, on one hand, I think as a teacher you should stay neutral, keep some distance to the content so to speak, but on the other hand, it still is good sometimes to steer the conversation a little. To soften those extreme opinions or to provide more contextual information.”
Similarly, Teacher 6 perceived a tension between her role to stimulate moral values and her role to create a safe discussion environment. More specifically, she argued that stimulating moral values might be counter-productive for students when they do not feel recognized by their teacher:
“I hope, that when the conversation is about homosexuality, students feel safe to say: ‘If my child was homosexual, he would not need to come home again.’ And that the teacher then does not immediately steer towards ‘this is against the Western norms and values’. That’s what I find hard, when you provide that student the feeling that his thoughts are wrong or unacceptable, you might give him even more reasons to stick to his ideas. Everyone wants to be recognized.”
