A case study on the use of a drama task to promote Dutch secondary school students’ historical imagination

Imagining what it was like to live in the past may help secondary school students to understand historical developments and situations. In this case study, the opportunities of a drama task are explored by using a mixed-method approach. In small groups, Dutch 14–15-year-old students examined historical sources and produced a short film clip on daily life in the Netherlands during the Cold War. Results indicated that both the students and their teacher perceived the drama task as motivating. The group discussions were rich in on-task utterances, and the students reported that they thought the task was valuable for gaining insight into thoughts and feelings of people in the past. However, the clips were relatively poor in information, and the assessment proved to be a challenge for the teacher.


Introduction
To understand historical events and developments it can be helpful for secondary school students to create an image of the past ( Barton and Levstik, 2004). Historical imagination can take many forms. For example, students can imagine the past with a focus on concrete details of a historical context (De Leur et al., 2019) resulting in a 'sense of period' (Dawson, 2009). A sense of period comprises concrete elements, such as information about daily life (such as clothing and housing) or ideas and attitudes (such as laws and religion). This might help students to construct a mental representation of the past (Collingwood, 1935;Huijgen et al., 2017). Moreover, historical imagination might focus on the thoughts and feelings of a historical actor. This activity is often referred to as historical empathy (Endacott and Brooks, 2018;Lévesque, 2008). Important questions related to historical empathy are how people would have experienced historical events, and how they would have felt during these events. This form of imagination can be source based, but students' own experiences also play a role in recognizing that some emotions, such as fear and love, can be seen as universal (Seixas and Morton, 2013). However, it is very difficult to recognize one's own focus on the present and abandon that focus for some time to look afresh at another time frame (Retz, 2015). Although too much involvement of students' own experiences can be seen as a threat to the historical plausibility of their image of the past, several scholars 'No,no,the Cold War was not that dramatic' 29 History Education Research Journal 18 (1) 2021 argue that historical empathy could contribute to students' historical imagination (for example, Brooks, 2009;Cunningham, 2009).

Constructing images of the past
Whether focused on the physical appearance of the historical context or on the thinking and actions of a historical actor, all imagination activities should be based on historical evidence (Lévesque, 2008). Collingwood (1935) argues that when confronted with knowledge gaps because of an absence of usable sources, one must try to fill these gaps 'like a detective' by using information from sources that are the most plausible. Thus, historical imagination is not the same as fantasizing about the past. However, when confronted with gaps in their knowledge, students have to imagine the things they do not know, using what seems reasonable from their own point of view (Barton and Levstik, 2004;Virta and Kouki, 2014). Therefore, scholars argue that these types of tasks can lead students towards misconceptions about the past, such as presentism, which is the transfer of values and information from the present to the past (Brooks, 2009;Wilschut, 2012). For the same reason, teachers can also be concerned about working with imagination tasks (Egan and Judson, 2008).
Despite these possible concerns, tasks in which students are asked to form an image of the past are used in everyday classroom practices (Cunningham, 2009). Historical imagination can be exercised and demonstrated in various ways (Fines, 2002). Especially when working with historical empathy, writing tasks are known to contribute to students' historical understanding (Brooks, 2008;De Leur et al., 2017;Virta and Kouki, 2014). A writing task focusing on imagining a concrete person can support historical understanding. In a study by De Leur et al. (2017), students reported that a writing task about a child in a factory supported them to picture the era of industrialization. Although in history textbooks writing tasks are omnipresent (De Leur et al., 2015), there are other ways for students to present their image of the past (Fines, 2002;Levstik and Barton, 2015;Munslow, 2016). For example, drawing tasks can evoke forms of historical imagination (De Leur et al., 2019;Dilek, 2010). Another promising way to promote historical imagination is a drama task (Rainer and Lewis, 2012). Drama activities can combine historical empathy with imagining concrete details of the past. When acting, students are forced to pretend that they are an actor from the past (historical empathy), and when they create a (role) play, the historical context (concrete details about the setting) has to be shown as well.

Beneficial effects of drama activities
Several scholars list the possible beneficial effects of drama activities on history learning. By impersonating a historical actor, drama can be a tool to deal with questions about human motivations, feelings and actions in history (Rainer and Lewis, 2012). A drama task can also help students to grasp the role of historical characters and to understand their actions (Egan and Judson, 2008;Stevens, 2015). Furthermore, drama could lead to an in-depth understanding of history through taking a detailed look at a particular time (Fennessey, 2000;Rainer and Lewis, 2012), especially when students themselves do the necessary research (Rantala et al., 2016). Moreover, drama tasks might enhance students' motivation, or their commitment to the lesson or topic they are studying. In an exploratory study on a scripted role play about the partition of India, Stevens (2015) conducted a questionnaire which shows that most students considered the role play engaging. Particularly, the emotional element appealed to students. Endacott and Pelekanos (2015) show that debating in the role of a historical character helps students to engage with the past. Wilhelm and Edminston (1998) argue that motivation can derive from the fact that, in a drama task, the students can take agency of the task and, therefore, the task becomes meaningful for them. Luff (2000) also reports that drama tasks can encourage student engagement because students themselves regulate the task, giving them a sense of ownership.

Designing and assessing drama activities
As shown, drama tasks might contribute to students' engagement with, and understanding of, historical events. In best-practice descriptions of drama activities, several scholars propose design principles for drama activities in the (history) classroom. McDaniel (2000) argues that to make a coherent story, a variety of sources should be provided, showing students that there is more to a topic than only the textbook. McDaniel (2000) and Harris and Foreman-Peck (2004) suggest that a task based on non-fictional individuals enhances students' active involvement in the task. Fennessey (2000) adds that details about humans in particular can encourage students to relate to the topic they study.
Moreover, a safe learning environment is essential for drama activities (Biddulph and Bright, 2003), although safety can also be a result of a drama activity (Stevens, 2015). Furthermore, a teacher must be comfortable with drama or role play (Luff, 2000). Some teachers prefer to discuss texts rather than working with drama tasks, because they think focusing on texts is more important (Stevens, 2015). In addition, some teachers lack confidence in their ability to create or manage a creative process (Newton and Newton, 2014;Pauw et al., 2018). Finally, there can also be a reluctance to use drama because it might hinder historical understanding. Too close (emotional) identification with the role a student is playing can result in less focus on the historical context (Endacott, 2014;Shemilt, 1984).
Because most drama tasks are open-ended, assessing such tasks is challenging (Rainer and Lewis, 2012). Kearney (2011) recommends the use of rubrics in order to include the full range of skills, processes and content goals in an assessment. Selwyn (1993) recommends using learner reports or self-evaluations. Nevertheless, both Selwyn and Kearney acknowledge that the difficulty of assessment can prevent teachers from using drama tasks in classrooms. De Leur et al. (2017) proposed to evaluate the processing of information from historical sources and prior knowledge, the use of concrete details, historical correctness, and the display of human thoughts and feelings when evaluating students' images of the past. Ashby and Lee (1987) propose a model with different stages of historical empathy, which was used in, for example, a study by Rantala et al. (2016), focusing on a simulation about the Civil War in Finland. However, Endacott and Brooks (2018) argue against the measurement of historical empathy. They stress that tasks related to historical empathy should be seen as a process towards historical understanding. Because of the personal factors contributing to historical empathy (that is, the personal context of the student), the learning process by means of historical empathy is highly individual. Furthermore, the follow-up after a drama activity is considered important (Luff, 2000;McDaniel, 2000). When discussing student products afterwards, there is an opportunity for elaborating on the content based on the activities of the students (Havekes et al., 2017).

Research question
There has been some research on drama activities in history education. However, empirical studies on the possible contributions of drama tasks to students' engagement 'No,no,the Cold War was not that dramatic' 31 in historical imagination, or what the produced images look like, are scarce. This case study aims to contribute to gaining more specific insights into the use of a filmed drama task by exploring the following research question: What are the learning opportunities in a drama task aiming at historical imagination in terms of students' engagement and the construction of an image of the past?

Method
Because this is an explorative study, aiming to describe the learning processes and outcomes, and the perspectives of the students while completing a drama task aiming at historical imagination, we performed a single case study with embedded cases (Cohen et al., 2007;Yin, 2009). We focused on one particular class and their teacher (the case) with five student groups (the embedded cases).

Participants
This study was conducted in a suburban, average-sized secondary school in the western part of the Netherlands. The history curriculum of this school is chronologically ordered, and mainly aimed at the acquisition of overview knowledge, as is the custom in the Netherlands (Van Straaten et al., 2018). All 29 students of one ninth grade group and their teacher participated. The students, aged 14-15, took their history classes at the Dutch intermediate stream preparing for higher education at a university of applied sciences (15 male and 14 female students). The students had some experience with the production of films, since they had completed a film task for a citizenship course one year prior to our study. The teacher of the class was a 26-year-old history teacher with a bachelor's degree in teaching history. She already had three years' experience of drama tasks in her lessons, both preparing small role plays and supervising a filmed drama task.

Drama task
For this study, we choose a drama task resulting in a filmed product (see Esslin, 1987). The topic that the students were studying was the Cold War. Therefore, the drama task focused on daily life in the Netherlands during the Cuban missile crisis in 1962. The task was designed in close cooperation by the teacher and the first author. The teacher set the goals for the task. She wanted to evaluate that students understood how the Cold War affected the daily life of a normal family, and she wanted to give students the opportunity to work with historical information in a creative manner. The teacher designed the rubric for the assessment. The first author assisted by collecting suitable sources and was present during the lessons, but did not intervene in the lessons or the assessment of the products.
The teacher divided the students into five groups of five or six students each, following the recommendations of Cooter and Chilcoat (1990) and Fischer and Frey (2012) that, in (dramatic) group work, six is the maximum number of students feasible for manageable discussions. The groups were mixed by gender, and each group consisted of both weak and strong performers in history class, according to their history grades. The students were asked to imagine an afternoon of a Dutch secondary school student aged 14 or 15 during the Cuban missile crisis, and to present their image in the form of a short film clip (2-4 minutes). The exact phrasing of the task was: 'Present a clip in which you make clear if and, if so, how, the Cold War affects your daily life.' To help the students, different sources were provided, such as a description of the events during the Cuban missile crisis; pictures of families and children playing, listening to the radio and doing household chores; and some suggestions for further reading on life in the Netherlands during the Cold War.
In the first 45-minute lesson of a series of five, the teacher delivered an interactive lecture about the Cold War. She talked about the end of the Second World War and the tension between the Soviet Union and the United States. She also discussed a picture of two children playing in a bomb shelter, and she displayed an original radio from the 1960s and a package of NATO emergency biscuits. With these concrete objects, she introduced the task. At the end of this first lesson, and during the following three lessons, the groups were allowed to work on the task. The teacher opened each lesson with a reminder of what the task was about and some admonitions about the group work, and she closed by urging the groups to fill out the group journals. During the last lesson, the clips that the students had produced were watched, and the task was briefly evaluated.
All the groups were free to decide among themselves who would do what and when, as long as they monitored the process in a journal. The teacher was available for all questions, whether on the subject of the Cold War or about the task itself.

Data collection
Because we wanted a detailed picture of the implementation of the drama task, we used various data. Furthermore, to increase the validity of the results, we used methodological triangulation (Miles and Huberman, 1994). Table 1 shows which data we used for which part of the analysis. We used the following data: (1) Audio recordings of the group work. During the group work, each of the groups had a voice recorder on their desk. By recording the group discussions, we tried to monitor the construction of a historical image as reflected in the clip.
(2) Classroom observations. During all the lessons, the first author was present in the classroom taking field notes on the students' task behaviour. (3) Group journals. We provided each group with a journal to be completed at the end of lessons 1, 2 and 3. In this way, we wanted to monitor the group process and the subject-related problems the groups encountered. The questions raised in the journals were: What did we do today? What do we want to do in the next lesson? What do we want to do at home? What did we want to know today, and how did we get to an answer? (4) Situational interest questionnaire. This questionnaire (Linnenbrink-Garcia et al., 2010) measures the students' interest provoked by a particular learning activity or lesson. It is a validated questionnaire consisting of 12 items scored by the students on a 1-6 Likert scale. Examples of the items are 'The task we just completed was fascinating' and 'I think that what I learned working on this task is useful'. We used this measurement to explore how students valued the task. The students completed the questionnaire at the beginning of lesson 5. (5) Learner report. With this learner report, we wanted to explore, on an individual level, what students took with them from the task (Van Kesteren, 1993). This report consisted of four sentences to be completed by the students: 'The first thing that comes to mind when I think of the task is…'; 'I have learned the following about the Cuban missile crisis…'; 'What I see before me when I think of daily life in 1962 is…'; 'What I will certainly remember from this task is…'. The students completed the learner report at the beginning of lesson 5. (6) Interview with the teacher. After lesson 5, we conducted a semi-structured interview with the teacher, focusing on her evaluation of the lessons, the group processes and the products the students handed in. (7) Interviews with students. After lesson 5, we interviewed triads of students from three groups in semi-structured interviews. We asked them about their perspectives on forming images in history education and let them evaluate their own clips. Due to time issues, it was not possible to interview Group 1 and Group 2. (8) Group product. The teacher assessed the products using a rubric she designed herself and that was known to the students. The rubric comprised a score for required elements (journal, recordings, finished product, turned in on time), task approach, use of the information provided, illustrative elements to demonstrate historical empathy/imagination, the historical correctness of the product, and a possibility for a bonus because of extraordinary creativity.

Data analysis
We mainly chose the group as the unit of analysis, only mentioning individual students when they stood out significantly. To analyse the data, we used 'pattern watching', a constant comparison method (Yin, 2009), visiting all the data several times to ensure internal consistency in the findings. First, we looked at students' engagement. To explore students' task behaviour, we focused on the degree of on-and off-task interaction and joint attention (Fischer and Frey, 2012). To better understand students' task behaviour and the situational interest they reported immediately after finishing the task, we explored the students' perceptions of the task. We focused on the students' positive and negative remarks in interviews and learner reports on what they valued in the task and what they thought was challenging. To explore the process, content and quality of students' historical imagination, we focused on elements relevant for historical imagination: the processing and use of the sources that were provided with the task, and the addition of both concrete details and human thoughts and feelings.
To explore the quality of the images, we used the assessment the teacher did herself, and her opinions about the quality of the drama products. The first author conducted the initial analysis, which was then discussed with the second author. We performed a member check on the teacher, and she agreed with our findings. Additionally, inspired by the guidelines for an audit procedure (Akkerman et al., 2008), a third history education researcher (the third author), who was not involved in the design of the study, did a check on the analysis of Group 3. This researcher was provided with all the data related to this group (recordings, student product and evaluation by the teacher, questionnaires, interviews). There was agreement about the interpretations. Our findings for all the groups are summarized in Table 2. The table is followed by a general description of the findings. Subsequently, to illustrate the whole production/learning process, we give a thick description of Group 3, which was selected because of the complete and rich data set.

Findings Students' engagement
The teacher reported in the interview that she had seen many on-task interactions in all the groups, but also that some students worked more seriously than others. This was confirmed by the field notes of the observer and the audio recordings. The students discussed the content of the sources, historical clothing and names, and the threat of a nuclear war. In all the groups, we saw students correcting each other and offering their own views on the sources or the content of the clip; for example, 'the problem was that neither Khrushchev nor America could win a war because in the war they could use nuclear bombs, and nuclear bombs kill everyone' (Group 1). In Group 5, one of the members sighed when the others suggested that back then no one dared to go out on the street: 'No, no the Cold War was not that dramatic.' The audio recordings and observations showed that the groups used different approaches in handling the task. Groups 1 and 5 focused on the form of the product (such as camera, props and acting); Group 3 initially focused on the content (sources, prior knowledge); Group 4 mainly focused on social interactions that were not really related to the task.
In most of the groups, the situational interest questionnaire scores were average (see Table 2), with a mean of 3.8 on a 6-point scale. However, there was quite some variation reported by individual students (SD = 0.87). In the learner reports, as well as in the interviews, most students confirmed that they enjoyed working on the task: 'We could talk and move, and we had freedom and responsibility.' The boy playing the lead part in the Group 5 clip reported: 'When I saw the clothes, I thought I didn't like it, but when I had everything on, I enjoyed it; I thought it was fun.' The students' motivation was also noticed by the teacher. She thought that the students appreciated 'No, no,    He joins her, and a mother presents them with cookies.
The alarm sounds, and the mother calls out: 'fast, to the shelter', and the three of them hurry outside.

9
Maximum points are given for the process, the content, and a bonus because of the level of creativity of the product.
*The recording of Group 1 lesson 2 failed, so we derive these findings from lessons 0, 1 and 3. the amount of freedom they had during the task since, she explained, normally they do not get much freedom during history class.
The learner reports showed that most students agreed that they had learned from the task -primarily that it had helped them to relate to the emotions of people in the past: 'Now I had to play or something. Then you have to truly mimic the feelings' (Group 5). In the interview, students from Group 5 elaborated on the difference between the task they had just completed and a 'regular' textbook task (reading a text and producing written answers): 'This is more difficult than writing because when you act out history, you have to make an image. When you complete a writing task from the textbook, you can copy the answers, but when you have to act, you learn things better.'

Construction of an image of the past
In all products, information from the sources was visible. The clips contained traditional role patterns (a mother who is cooking and cleaning), the threat of a nuclear bomb is elaborated on, and Americans and Soviets are mentioned. Some groups tried to use old-fashioned names for their characters. Additionally, some students were dressed in a way that they thought was the custom sixty years ago, thus adding concrete details that did not derive from the sources. In all clips, human emotions were shown, mainly fear and insecurity.
In the learner reports and the interviews, students provided us with insight into the understanding gained from the task. Group 4 told in the interview: 'Making the clip, we now know more about those people. Because we had to play their emotions, which sounds weird, but imagining that period you could imagine their feelings a little.' The students elaborated: 'For example, people lived in fear. However, they still went on with their lives. They prepared slightly with shelters. They hardly had any electronics, and their clothes were different.' The teacher noted that 'This class knows a lot about daily life and human details, but has hardly any knowledge of politics; another class that did a traditional pencil-and-paper task learned more about politics, but has no idea about daily life.' The assessment of the quality of the images constructed by the students proved challenging. The teacher thought that the actual products were disappointing, because she expected the clips to be richer in information. However, the recordings of the group talks and the interviews afterwards showed that the students' conversations included much more historical content than they made visible in the product. In the interview, Group 3 explained: 'Our clip is now about school and a bomb drill, so the teacher cannot know what we learned about other stuff. But we did.'

Group 3
To examine the learning processes and experiences of the students in more detail, we will present a thick description of Group 3. We will follow Group 3 during the consecutive lessons, and then describe their perceptions of the task, as shown afterwards in the interview.
Group 3 consisted of two boys and three girls, one of whom was absent during the final lesson. In the ten minutes the students were given for consultation during the first lesson, they started right away by discussing the task and the clothes they could use in their product. During the second lesson, Group 3 divided the information provided, looked at some of the internet links provided (one of which, an information clip about the Cuban missile crisis made by Dutch public school television, appeared in the actual product), and started studying them. The journals showed that the 'No, no, the Cold War was not that dramatic' 39 students summarized each source and wrote down which elements from the sources they could possibly use for their product, such as, 'Source 1 mentions people being scared and talking to each other, so we could use people gossiping in our clip.' The students constantly communicated about their associations, such as when discussing the source about television: 'Oh no, they watched television with a whole group? Yes, on a small screen. Kind of like a microwave.' This was later followed by an opinion: 'How funny, there were only children's programmes on Sunday.' Additionally, the group discussed the Cold War in general: 'It was about the threat of Russia and America. And the missiles on Cuba.' The conversation of the third lesson was almost entirely dedicated to deciding on a plot: 'We will do children who are walking home while discussing that war, and then, they watch the news or something … Did they have TV? No. Or did they have radios? Yes, radios.' The students talked about setting, clothing and hair, and they made a list of props they might need ('torch, radio, canned foods'). They also discussed how to make the clip look authentic: 'Look, if we just make the clip in black and white, then you do not notice the modern stuff that much.' During the fourth lesson, the students did the actual filming. At times, the students discussed some issue that was related to creating a historically correct image: 'Look, T. was smart, she removed her watch.' However, most of the talking was about practical problems: 'No way am I going to lie down on the grass!' In the learner reports, students stated that they would remember that this task was 'a nice way to learn history', and they stressed the fear and nervousness people must have felt during the Cuban missile crisis.
In the group interview, the students explained that, because they had to make a clip, they were forced to construct an image. To do that, they had to think about the information provided. In some cases, this enhanced their reflection on the differences between the past and present situations: 'We studied a lot of information, and then you have an idea, but, well … now, we are free naturally, but then, every time there was fear that something could happen. But you have to be able to imagine that.' Additionally, when shooting the clip, they had to think of issues related to a historical image: 'You see, the cars that are out there today, they were not there then.' According to the students, the task has primarily helped them to relate better to the emotions of the people from the past compared to a regular pencil-and-paper task: 'When you read the textbook, you think "that country does this or that", and then, I think "whatever". But, when you study information to use in our own clip, then I can imagine how people experienced their life.' They explained: 'When reading a textbook, sometimes you feel sad, but when you have to make a clip yourself, you feel more, because you have to think how people must have felt.' Although the students from Group 3 said that they were quite satisfied with their own product, they agreed that the actual clip did not reflect everything that they had learned during the preparation of the clip. They said, 'You see, we brought a first aid kit, but we did not use it', and, 'We talked about how to prepare for a nuclear crisis with extra food at home, but we haven't put that in the clip. But still, we learned about that.' The students discussed many historical issues during the lessons in preparing to film the clip, but most of what was discussed was not included in the final product. For example, the radio, the workload of the mothers, or even the fear that the students attributed to people in 1962 are not present in the clip. On the other hand, the main historical element that the students did include in the clip (a bomb drill, see Table 2) is not shown correctly, because the students first drop on the ground, and then go inside and hide under a desk. Apparently, the students blended the information from the source about what to do when you are outside and what to do when you are inside.

Conclusions and discussion
Our research question was: What are the learning opportunities in a drama task aiming at historical imagination in terms of students' engagement and the construction of an image of the past? We have seen that, in terms of engagement with historical imagination, the drama task seems to elicit students' engagement, focus on source work and interest. However, the image of the past that the students present in their clips does not fully reflect their group talk. Historical imagination is mostly present in the form of some historical empathy.
Regarding students' engagement, we have seen both on-and off-topic conversation. The students discussed relevant information from the sources, details about hair and clothing, and the possible plot of their clip. The students report moderate situational interest. According to the observations, the teacher and the students themselves, there was a safe enough learning environment.
Our findings regarding interest and motivation are in line with Wilhelm and Edminston (1998). Both the students and the teacher expressed their involvement in the task. This was also seen in the observations: the students were committed to the task. The teacher suggested that the fact that the students could plan their own work and had choices in what they wanted to show in their final product motivated them. However, some groups had quite a lot of 'downtime', which they spent talking about subjects other than the Cold war. This can easily be addressed by additional guidance from the teacher. We will elaborate on that below. Additionally, the actual making of the clip proved to be motivational. In the interviews, the students explained that they liked to create an image because, they say, in regular tasks, they just copy from the textbook. In prior research, students reported similar thoughts (De Leur et al., 2019;Yeager and Doppen, 2001): when required to write a text, you just copy, when required to make an image, you start wondering. That process of imagination leads, in several cases, to comparing the past with the present, which is also a factor that is considered motivational (Van Straaten et al., 2018).
Regarding the image of the past that the students constructed, we have seen that the filmed drama task proved to be an incentive for historical imagination and an opportunity for historical thinking. The groups processed the information from the sources, and, in one case, the students actively searched for additional information. Although in the products, most groups focused on bomb drills, in the learner reports and audio recordings of the group talk, all the other information (for example, about households and the media) can be identified. There are no major historical faults in the clips. Obviously, there were some inaccurate details, such as modern cars when filming was outside, and some students wear sneakers underneath their long skirts, but the students themselves were well aware that these details were incorrect, and sometimes they had no other choice. Differences between 'back then' and now were noticed, both during the lessons and in the interviews. Some students tried to compare the situation during the Cold War with their present freedom and wealth. Although this comparison can lead to forms of presentism, which of course should be addressed by the teacher when they arise, comparing the past with the present can also be an indication that students are engaged in historical reasoning, an important feature of history learning (Van Boxtel and Van Drie, 2018).
The teacher reported that the task was valuable because the students now have an image of daily life in the 1960s, whereas with a regular textbook task, they would have known only about politics. However, she had hoped that the students would learn both about daily life and about politics when working on the drama task. She was a little disappointed that the students focused only on daily life, when sources on politics were also available to them. This seems a disadvantage of this drama task: by focusing on acting out a day in the life of a child in the Netherlands, it was difficult for them to integrate information about political events and to acknowledge the big picture of the Cold War. Thus, when this (also) is the aim of the teacher, the task itself and the criteria to assess the product should be phrased differently.
The students themselves reported in the interviews that they thought they came closer to the past by imagining the thoughts and feelings of the people, which they found valuable. This is in line with findings of Endacott and Pelekanos (2015). Although the word 'emotions' did not appear in the task itself, the students report extensively on emotions. The main emotion that they attribute to their protagonists is fear, which is not very surprising since the project started with an introductory lecture based on a photograph of two children in a shelter. The focus on fear can be seen as a poor result, because the image of scared people is easy and straightforward, but does not do much justice to the complexity of the historical situation and the experiences of people in the past. This finding is in line with, for example, Arnold (1998) and Shemilt (1984), who warn that too much identification with a historical actor can result in poor historical understanding. However, two groups managed to nuance the image of fearful people by stressing that daily life went on as normal. This was not obvious in the sources and, thus, was a product of their historical imagination: 'How would it possibly have been?'.
Previous research suggests that students who are (in a writing task) invited to imagine themselves in the past tend to judge (actions of people in) the past (De Leur et al., 2017). In this study, we did not note such judgements in the products. The students seemed to have tried faithfully to construct an image of the past. However, the topic of the clip, daily life in the Cold War, may not encourage judgements because there is no dilemma or action that can be disputed in the task.
The quality of the students' images proved difficult to assess. The group talks were richer in information than the clips, so the clips seemed to represent the learning process less than was hoped. After watching the products, the teacher was disappointed, and she regretted not having prescribed some compulsory elements. In hindsight, she thought that the task was too open, and she would make the task more specific next time, allowing less freedom regarding the content of the clips. She would have liked to know for sure which pieces of information would be in the clips. When using the rubric she herself designed, the teacher was able to reward students who worked fruitfully, even when their product was not very good. She thought that in this way she could do justice to the learning process as a whole, whereas only assessing the finished product would not have been a fair assessment. Nevertheless, one can consider whether this type of task is suitable to be (summatively) assessed at all. Perhaps a formative assessment addresses the creative process of a drama task better. It would be interesting to explore to what extent a drama task can be used as a tool or an instrument while learning history. This is in line with the proposal of Endacott and Brooks (2018) to see the product of an activity based on historical empathy as a step on the path towards historical understanding. Endacott and Brooks (2018) therefore argue for a focus on reflection activities instead of summative assessment.
Although teachers and students can be uncomfortable with a drama task (Luff, 2000;Stevens, 2015), we did not encounter this challenge. A possible explanation is that the teacher participating in this study was already experienced in working with drama and film tasks, and the students had already produced a short film. One student expressed that she did not like the clips being watched by the whole class. This can be accounted for by the lack of debriefing. Havekes et al. (2017) stress that, with this kind of open task, the teacher-led classroom discussion is very important, and Luff (2000) suggests that some form of debriefing should be a part of every drama task.
This study has several limitations. Although we have no evidence that the students felt restricted by the voice recorders on their desks or because of the presence of the researcher in the classroom, it is possible that they altered their behaviour because of the research setting. The teacher, on the other hand, did report feeling restricted because of the research setting; she would have liked to interfere more, but was afraid to influence the students too much. More guidance from the teacher might improve the quality of the students' talk and the students' products, and may reduce the amount of time students spend talking off-topic. Second, because the task focused on the production of an image of daily life, we do not know whether, and, if so, how, this type of task promotes historical thinking and reasoning about the Cold War as a whole. It would be interesting to further explore the possible benefits of creating an (initial) historical image when learning about complex historical situations or developments, and how to embed these types of tasks in the curriculum. Finally, because this is a case study, we cannot generalize the findings. However, we do see promising topics suitable for further research: the applicability of drama tasks in different age categories or with different content, constructing design principles for both task and assessment instruments, and exploring the possibilities of (group) discussions about the drama product, along the way and afterwards, in whole-class discussions.
The findings illustrate that teachers need to design drama tasks carefully and select sources which give information about the topics they want to see processed by the students. Teachers have to choose to what extent they would like to prescribe topics that students must include in their image. Compulsory topics ease the assessment of the image. However, more freedom for the students might result in more motivation. Furthermore, it can be recommended to include in the assessment procedure both the process and the product, recognizing that the group discussions that lead to the production of the clip are also a learning opportunity. Although the final products may vary widely in quality, the production of the clip is important for the students. Having to produce a clip stimulates them to imagine the life of people in the past. Finally, it is important to debrief the task in a classroom discussion or through a written report.
Despite the fact that implementing drama tasks in history education might be challenging for history teachers, this study shows many positive possibilities for using drama tasks to promote students' understanding of past events.

Notes on the contributors
Tessa de Leur is a teacher-trainer at the Amsterdam University of Applied Sciences, the Netherlands. Her research focuses on tasks that promote historical imagination.
Carla van Boxtel is Professor of History Education at the University of Amsterdam, the Netherlands. Her research focuses on the teaching and learning of history in and outside school. She has published widely about the teaching and learning of historical reasoning.
Tim Huijgen is an assistant professor and a history teacher educator at the University of Groningen, the Netherlands. His research focuses on domain-specific learning, especially the teaching and learning of historical reasoning.