Toward Historical Perspective Taking: Students’ Reasoning When Contextualizing the Actions of People in the Past

: An important goal of history education is to promote the student’s ability to perform historical perspective taking (HPT). HPT refers to the ability to understand how people in the past viewed their world at various times and in various places to explain why they did what they did. In this study, we assessed a sample of 15-and 16-year-old pre-university students ( n = 170) to determine their ability to contextualize the actions of people in the past. Subsequently, we explored their reasoning ( n = 36) to uncover their contextualization process. The results of this mixed methodology study indicate that most of the students in the sample performed well when engaging in HPT. Moreover, protocol analysis identiﬁed the different reasoning strategies that students employed to successfully perform HPT. The results of this study provide insight into history instruction regarding HPT and into strategies for designing valid and reliable HPT tasks.

In his book Logics of History, social historian William Sewell, Jr. (2005) noted that historians should respect the differences that separate one period from another. He argued, "We cannot know what an act or utterance means and what its consequences might be without knowing the semantics, the technologies, the conventions-in brief, the logics-that characterize the world in which the action takes place" (p. 10). Other historians also stress the importance of considering the contextual circumstances when interpreting historical phenomena (e.g., Bevir, 2002;Gaddis, 2002;Tully, 1988). Accordingly, as student ability to contextualize historical phenomena is considered an important component of historical thinking, such conceptualization is being incorporated into history education worldwide (e.g., Lévesque, 2008;Seixas & Morton, 2013;Van Drie & Van Boxtel, 2008). In history education, it is possible to contextualize historical sources and phenomena, including persons, events, and developments (Havekes, Coppen, Luttenberg, & Van Boxtel, 2012). When history education researchers discuss the contextualization of the actions of people and groups in the past, they often use the term historical perspective taking (HPT; e.g., Davis, Yeager, & Foster, 2001;Doppen, 2000). 1 Though people in the past lived under different circumstances and viewed the world through different belief systems, many students might assume that people of the past had the same goals, intentions, attitudes, and beliefs as people in today's society, and as such, this presentism might result in misunderstandings about the past (Barton & Levstik, 2004;Lee & Ashby, 2001). For example, without the ability to perform HPT, students could not explain that Julius Caesar could not have breakfasted in Rome and dined in the Gaul region of France on the same day, as the transportation necessary for such a trip was not available during Caesar's time (Lévesque, 2008). Engaging in HPT could avoid presentism and help students understand and explain historical agents' decisions and historical phenomena (Van Boxtel & Van Drie, 2012). Some scholars also argue that HPT could contribute to citizenship in multicultural societies as it promotes the recognition and understanding of other people's views (e.g., Barton, 2012;Den Heyer, 2003;Rüsen, 2004). For example, Seixas and Peck (2004) argued that to promote students' social and political orientation and moral judgment, they must engage in HPT assignments.
Despite the importance of HPT in enhancing students' historical thinking and promoting citizenship among students, recent research has indicated that students may struggle when asked to perform thinking skills, such as HPT (e.g., Beyer, 2008;Huijgen, Van Boxtel, Van de Grift, & Holthuis, 2014;Reisman & Wineburg, 2008;Van Boxtel & Van Drie, 2012), and that history teachers may lack the requisite knowledge to promote historical reasoning competencies, such as HPT (e.g., Achinstein & Fogo, 2015;Bain & Mirel, 2006;Grant & Gradwell, 2010). Moreover, valid assignments and measurement instruments to assess students' historical reasoning competencies, such as HPT, are scarce (Breakstone, 2014;Reich, 2009;Rothstein, 2004;VanSledright, 2013). Therefore, to understand how students learn history and how they improve as a result of such learning, more information is needed regarding how students reason when performing historical reasoning competency tasks and regarding the development of instruments that operationalize this type of reasoning (Hartmann & Hasselhorn, 2008;Huijgen et al., 2014).
In this study, which uses an HPT instrument developed by Hartmann and Hasselhorn (2008), the ability of 15-and 16-year-old pre-university students (n = 170) to contextualize the actions of people in the past was assessed. Furthermore, we explored, using thinking-aloud methodology, how a sample of 15-and 16-year-old pre-university students (n = 36) reasoned to uncover their contextualization process when working with the HPT instrument. The results of this study provide insights into the difficulties students experience when engaging in HPT and into the validity and reliability of HPT classroom assignments, thereby helping teachers to promote their students' ability to perform HPT.

HPT: A Conceptualization
Because of the critical role HPT plays in students' understanding of history and in promoting the competencies students need to successfully participate in civic life, the ability to perform HPT is incorporated into the formal K-12 history curricula of, for example, the United Kingdom (Cooper & Chapman, 2009;Department for Education, 2013), Australia (Australian Curriculum, Assessment and Reporting Authority, 2014), Canada (Peck & Seixas, 2008), Germany (Hartmann & Hasselhorn, 2008), Finland (Rantala, 2011), Belgium (Wils & Verschaffel, 2012), and the Netherlands (Van Boxtel & Grever, 2011). Though in many states of the United States, HPT and similar reasoning competencies have appeared to play only a marginal role in the formal curricula (e.g., Evans, 2011;VanSledright, 2008;Wineburg, 2001), with the recent development of The College, Career, and Civic Life (C3) Framework for Social Studies State Standards (National Council for the Social Studies, 2013), more attention, in the near future, may be given to implementing reasoning competencies, such as HPT, in state curricula. For example, two objectives of the C3 Framework are that, by the end of grade 12, students will be able to "analyze complex and interacting factors that influenced the perspectives of people during different historical eras" and "analyze how historical contexts shaped and continue to shape people's perspectives" (p. 47).
In the literature, different definitions of HPT exist. For example, Seixas and Morton (2013) defined HPT as an attempt to see through the eyes of people who lived in other times and circumstances that are sometimes far removed from our present-day lives. Levstik (2001) defined HPT as the ability to see how people acted in the past and understand why they acted as they did. To achieve HPT, scholars stress the importance of understanding the social, cultural, intellectual, and emotional settings that shaped people's lives and actions (e.g., Lee & Ashby, 2001;Seixas & Peck, 2004). Moreover, it must be emphasized that knowledge and understanding of chronology are important for successful HPT . Accordingly, HPT is a complex historical reasoning competency that consists of several components.
From the extant literature, we identify three interrelated components needed to successfully perform HPT. These include applying the awareness that a present-oriented perspective might hinder the understanding of people's actions in the past, demonstrating historical empathy, and reconstructing an adequate historical context. The first component is to be aware of a possible present-oriented perspective and the consequences of this perspective when examining the past. Present-oriented thinking, or presentism, is the bias by which people assume that the same goals, intentions, attitudes, and beliefs that exist in the present day existed in the past (Barton & Levstik, 2004). Forms of displaying a present-oriented perspective include viewing people in the past as stupid or assuming that people in the past had the same knowledge available to them that we currently have (Lee & Ashby, 2001;Shemilt, 1983). This perspective could cause misconceptions that lead to incorrect conclusions about the past and thus hinder successful HPT (Reisman & Wineburg, 2008). Although we can never be perfectly non-presentist (e.g., VanSledright, 2001;Wineburg, 2001), students must understand that the past differs from the present when interpreting historical phenomena and the decisions of historical agents (Seixas, 1996;Seixas & Morton, 2013). Students who are aware of the difference between the past and the present and acknowledge their present-oriented perspective might demonstrate this awareness by explaining that people in the past did not know what we now know or that people thought differently in the past.
The second component is to exhibit historical empathy. Historical empathy refers to placing oneself in the position of people in the past to understand their motives and values regarding their decisions and actions (e.g., Cunningham, 2009;Endacott & Sturtz, 2014). Although some scholars have argued that historical empathy can never be fully achieved and is idealistic because it is impossible to put oneself in the shoes of a historical agent (e.g., Kitson, Husbands, & Steward, 2011;Riley, 1998;Wineburg, 1998), many scholars have concluded that historical empathy contributes to insights about historical agents' decisions (e.g., Brooks, 2011;Endacott & Brooks, 2013;Kohlmeier, 2006). However, though history education research has debated the extent to which historical empathy is an affective or cognitive achievement (e.g., Virja & Kouki, 2014), we consider historical empathy as a combination of affective and cognitive processes, following the conceptualization of scholars such as Endacott and Brooks (2013). It is further posited that connecting with known and familiar emotions of people in the past as an affective process might promote historical empathy and understanding of historical agents' decisions (Riley, 1998;Skolnick, Dulberg, & Maestre, 2004). Furthermore, considering the roles and positions of different historical agents in society and how such positions may have affected their views on historical phenomena as a more cognitive process could also contribute to historical empathy and to the understanding of historical agents' actions (Bermúdez & Jaramillo, 2001). In this study, we use the concept of historical empathy as putting oneself in the shoes of a historical person by considering his or her emotions, role, and position. For analytical reasons, we consider the reconstruction of a historical context as a distinct component.
The third component is the reconstruction of the historical context. Foster and Yeager (2001) argued that students must possess historical context knowledge, which includes knowledge about chronology, before they can interpret historical phenomena and historical agents' actions. To reconstruct a historical context, students can use different frames of reference, including a chronological frame of reference, a spatial frame of reference, or a social frame of reference. The chronological frame (e.g., Dawson, 2009;Wilschut, 2012) includes knowledge about the time and the period as well as the sequence of significant events and developments. For example, when attempting to understand why people in Germany in the 1930s voted for the Nazi Party, it is important to know the sequence of the First World War, the economic crisis of 1929, and the rise of Hitler. In contrast, the spatial frame focuses on knowledge about geographic locations and scale (e.g., De Keyser & Vandepitte, 1998;Havekes et al., 2012), such as knowledge of where Germany is located in Europe, what countries share boundaries with Germany, and what countries are near Germany. The social frame includes not only knowledge about human behavior and the social conditions of life but also knowledge about social-economic, social-cultural and social-political developments (e.g., Pontecorvo & Girardet, 1993;Shemilt, 2009;Van Boxtel & Van Drie, 2012), such as knowledge of the poor German economic circumstances and the anger Germans had regarding the Treaty of Versailles. Some studies (e.g., Berti, Baldin, & Toneatti, 2009) contended that if students do not possess sufficient knowledge to reconstruct a historical context, they may use historical empathy (by referring more to specific characteristics of the historical agent) to perform HPT.

Students' Ability to Perform HPT
Are secondary school students cognitively capable of taking a historical perspective? Using short historical stories and classifying students' answers to questions related to those stories, Hallam (1970) and Kennedy (1983) concluded that students under the age of 16 years lack historical reasoning competencies, such as HPT. Compared with adults, elementary and secondary school students do indeed experience greater difficulty taking another person's perspective, particularly when that other person does not possess the same knowledge that they have (Bloom & German, 2000;Perner, 1991;Wellman, Cross, & Watson, 2001). Birch and Bloom (2007) discussed the "curse of knowledge," which is a cognitive bias that makes it difficult for students who have more knowledge to think from the perspective of lesser-informed people. This inability hinders the successful implementation of HPT in history education, as students must be aware that much of the information and knowledge they possess was not available to people in the past.
However, studies on students' ability to perform HPT have shown that even upper elementary school students are capable of some form of HPT and can overcome tendencies of presentism (e.g., Barton, 1997;Davis et al., 2001;Foster & Yeager, 1999;VanSledright, 2002). In their Concepts of History and Teacher Approaches 7 to 14 (CHATA) project, Lee, Dickinson, and Ashby (1997) examined how students between the ages of 7-14 understand the nature and status of different historical claims. They found that some students between the ages of 11-14 were beginning to distinguish between what they know and what the historical agent knew at that time. Berti et al. (2009) interviewed a total of 150 students aged 8-25 years about the concept of the ordeal during the Middle Ages and concluded that nearly every student understood that the ordeal involved the intervention of God and was related to religious beliefs that differ from the beliefs held in the present. Hartmann and Hasselhorn (2008) investigated how 170 German 10th-graders (mean age of 16) performed on an HPT instrument. They found that approximately 90% of the participants in the study successfully performed HPT. Huijgen et al. (2014) used the same instrument to test the ability of 1,270 elementary and secondary students aged between 10-17 years to perform HPT. Their results showed that even upper elementary school students are capable of performing some elements of HPT, though older students performed HPT more successfully than younger students.

Task Approaches and the Ability to Perform HPT
Research has indicated that not only domain-specific knowledge, understanding, and strategies are important for solving problems, but also that more generic task approaches are important, such as carefully analyzing a problem and evaluating decisions (e.g., Alexander, 2003;Bransford, Brown, & Cocking, 2000). Only a few studies have focused on the use of task approaches in combination with contextualizing historical sources and historical agents' actions. When investigating how students contextualize and date historical images and documents, Van Boxtel and Van Drie (2012) found that students who rushed to a conclusion or ignored information regarding the source more often failed to contextualize the source compared with students who approached the task systematically and used many clues provided by the source to generate alternative hypotheses. Wineburg (1998) investigated how two historians created a historical context from a historical text noting that specification of ignorance could promote the ability to create an adequate historical context. This specification of ignorance can refer to expressing puzzlement, asking questions, or specifying gaps in knowledge.
Though research has been conducted on how certain tasks, such as class discussions (Kohlmeier, 2006), source work (Brooks, 2011), and writing assignments (Brooks, 2008), can support components of HPT, important questions regarding the process of HPT and the difficulties students experience when performing HPT remain. For example, do students who rush to conclusions or who do not display their specification of ignorance perform more poorly on HPT than students who express doubt, ask questions, and understand the consequences of what they do not know?

Research Questions
Teachers, educators, and researchers are still missing relevant information about why some students successfully perform HPT while others fail. In this study, we answer the call of previous research that argues for the use of think-aloud methods to identify students' reasoning when performing HPT and to further validate instruments that assess students' ability to perform HPT. We therefore specify the following two research questions: 1. What are the HPT abilities of 15-and 16-year-old pre-university students? 2. How do 15-and 16-year-old pre-university students reason when completing an HPT instrument?

Research Design
To answer our research questions, we used a mixed-method research design incorporating an HPT instrument as a student task. First, we conducted quantitative research to examine 15-to 16-year-old pre-university students' general level of ability to perform HPT. Next, we conducted qualitative research using the think-aloud methodology to explore students' underlying reasoning processes when performing HPT. In other words, we investigated how these students solve the assignment of the HPT instrument. The thinkaloud methodology, which has been widely used to capture students' reasoning processes (Van Someren, Barnard, & Sandberg, 1994), is used as surveys and experiments would be unable to provide the rich and deep information about students' reasoning processes that is necessary to answer our research questions (Creswell, 2009;Macpherson, Brooker, & Ainsworth, 2000).
We chose a mixed-method design because combining quantitative and qualitative research provides a better understanding of a research problem or issue than does the use of either research approach alone (Creswell & Plano Clark, 2011;Tashakkori & Teddlie, 2010). Moreover, we focused on pre-university students aged 15-16 because, based on previous research (e.g., Berti et al., 2009;Hartmann & Hasselhorn, 2008;Huijgen et al., 2014), we concluded that students are able to perform HPT at this age, thus enabling us to investigate the reasoning that underlies one's ability to take a historical perspective.

The HPT Instrument
An HPT instrument developed by Hartmann and Hasselhorn (2008) and translated into Dutch by Huijgen et al. (2014) was selected, as this instrument is suitable for research on a large group of students and refers to a historical topic that has been taught to the students participating in this study, thus resulting in sufficient prior knowledge. The HPT instrument consists of a hypothetical scenario referring to the rise of the Nazi Party in Germany in the 1930s. The central historical agent in the scenario is a young man (Hannes) who struggles to decide which political party to vote for in the next election. An authentic historical source was not included in the instrument because Hartmann and Hasselhorn did not want to conflate students' HPT ability with their ability to understand historical sources. The students' central assignment was to decide if Hannes is willing to vote for the Nazi Party.
In relation to the scenario, Hartmann and Hasselhorn (2008) formulated nine items that corresponded to three categories: the present-oriented perspective (POP), the role of the historical agent (ROA), and historical contextualization (CONT). These three POP items may trigger possible forms of presentism in the students. For example, the first item, "He definitely will not vote for the NSDAP [National Socialist German Workers' Party, or Nazi Party]. No one approves of what this party has done to the world," illustrates knowledge that contemporary society possesses, but the German people living in 1930 did not possess this level of knowledge regarding the Nazi regime. This category aligns with our first conceptualized component of HPT, specifically, applying awareness that a present-oriented perspective might hinder the understanding of people's actions in the past. The three ROA items refer to the historical agents' personal situation, such as the agents' family life. For example, the item "Because his father's business is almost bankrupt, he might vote for a party that protects small business owners" may trigger possible affective connections between the students and the historical agent through, for instance, recognizable emotions, such as protecting family members, thus aligning with our conceptualized affective processes of historical empathy, or trigger considerations of the position of Hannes' family in society, such as wealth and influence, thus aligning with our conceptualized cognitive processes of historical empathy. In contrast, the three CONT items display historical contextualization and form the opposite of the POP items. For example, the item "Hannes has little experience with democracy. He probably does not know the risks associated with the NSDAP and thus will probably vote for the NSDAP" should trigger the reconstruction of the social-political context of Germany in the 1930s. In this scenario, students would have to know that Germany was an empire led by one strong leader for a long time and that the German people may want to return to this state, in which case, they would view Hitler as the new strong emperor. The CONT category aligns with our conceptualization of the third HPT component, namely, reconstructing the historical context. Hartmann and Hasselhorn (2008) tested their instrument among 170 German 10th-graders (mean age of 16). In a confirmatory factor analysis, they found that the POP and CONT items constituted one factor and that the two ROA items constituted the second factor. One item in this category (ROA1) displayed loadings above 0.40 on both factors and was excluded from further analysis. Huijgen et al. (2014) translated the instrument into Dutch and tested 1,270 Dutch upper elementary and secondary school students, ranging in age from 10-17 years. Their confirmatory factor analysis also indicated that the POP items and CONT items constituted one factor and that the three ROA items pertaining to the role of the historical agent constituted the second factor. In contrast to Hartmann and Hasselhorn's finding, the item ROA1 did not violate the simple structure. To assess the instrument's face validity, Huijgen et al. (2014) asked ten expert history teachers to sort the nine items on the instrument into the three categories. To determine the level of agreement among these experts, the authors calculated the Fleiss kappa, which at 0.64 indicated substantial agreement (Landis & Koch, 1977). Additionally, Hartmann and Hasselhorn found good inter-coder consistency (κ = 0.83) when four coders sorted the items into the three categories (POP, ROA, and CONT).
To calculate an HPT score, we used the same scoring system and 4-point scale as Hartmann and Hasselhorn (2008) and Huijgen et al. (2014). The selection of responses in the instrument's first column of the POP items ("Does not fit his situation at all"; also see the Appendix) receives 4 points. Second column responses receive 3 points, third column responses are awarded 2 points, and fourth column responses receive 1 point. The role of the historical agent and historical contextualization items had the opposite coding system, i.e., from left to right, as these items reflected good HPT ability. Selecting first column responses yield 1 point, second column responses receive 2 points, third column responses are awarded 3 points, and fourth column responses receive 4 points. A mean category score was calculated by summing the scores of the items in the category and dividing the total by three (because each category has three items). A total mean HPT score was calculated by adding the different mean category scores and dividing this score by three (because the instrument has three categories). Mean HPT scores < 2.50 denote inadequate ability to perform HPT. The 2.50 limit was chosen because it is the middle of the instrument's 4-point scale. Mean HPT scores ≥ 2.50 < 3.00 denote adequate ability to perform HPT, scores ≥ 3.00 < 3.50 denote good ability, and scores ≥ 3.50 denote excellent ability.
To test students' prior chronological knowledge about the historical topic, we included four multiple-choice items in the instrument (see Appendix). These items focused on important German historical events for the period 1900-1950, such as the year of the great worldwide economic depression and the year Hitler came to power in Germany. We did not ask for more detailed knowledge about significant events and developments during this period because we did not want to reference too much topic knowledge before students were asked to complete the HPT instrument. The four historical events presented in the questions were chosen because of their importance and their relationship to the scenario in the instrument. Each correct answer to a question yielded 1 point, resulting in a maximum score of 4.

Research Context
In the Netherlands, all children receive elementary education between the ages of 4-12 years. They receive education in, for example, writing, reading, geography, history, math, and English. Around age 12, the children transition to secondary education. This is when the first differentiation among three educational levels occurs. Approximately 60% of the students go to pre-vocational schools (duration of 4 years), 20% receive general higher secondary education (duration of 5 years), and 20% receive pre-university education (duration of six years). The determination of a student's level of education is based on the advice of the elementary school and supported by a mandatory standardized test that measures the student's attainment of certain standards (e.g., language, world orientation, mathematics) in elementary education. Only a pre-university degree allows access to Dutch universities. Furthermore, the educational quality of all elementary and secondary schools is monitored by the Dutch Inspection of Education.
The ability to perform HPT is included only in the formal history exam program of general higher secondary education and pre-university education. A total number of 323,498 students attend upper secondary education between the ages of 15-18, of which 49% are placed in higher secondary education and 51% receive pre-university education. The gender distribution for general higher secondary education is 49% male and 51% female, and for preuniversity education it is 47% and 53%, respectively (Statistics Netherlands, 2014).

Quantitative Research Sample and Data Analysis
Using the HPT instrument, we tested 174 10th-grade pre-university students from seven schools (four urban, three rural) to examine their ability to perform HPT. Four cases were excluded due to missing data, leaving us with data from 170 students for further analysis. The mean student age was 15.1 years, and the gender distribution of the sample was 54% female and 46% male. The participating schools generally matched the total population in terms of student enrollment and graduation rates (Statistics Netherlands, 2014). History was a compulsory subject for all 170 students, and students received two history lessons, each ranging from 50-60 minutes per week.
Approximately one year prior to the study, the students had studied the history of Germany. The foci of the course included the First World War, the rise of Hitler, the Second World War, and Germany's role in the Cold War. To examine how the 170 students performed, we calculated students' mean HPT score, mean category scores (POP, ROA, and CONT), and mean prior knowledge scores.

Qualitative Research Sample and Data Analysis
For our qualitative research sample, we used non-probability sampling to select ten history teachers from ten schools, five urban and five rural. The participating schools generally matched the total population in terms of student enrollment and graduation rates (Statistics Netherlands, 2014). We asked the ten selected teachers to randomly select four 10th-grade pre-university students. Four students did not agree to participate in the study, resulting in a sample of 36 students. In this sample, the mean student age was 15.6 years, and the gender distribution was 19 female students (53%) and 17 male students (47%). The students' answers were videotaped and transcribed for further data analysis (Ericsson & Simon, 1993). The interviewer was instructed to encourage students to think aloud and to read the instrument's items aloud to trigger students' reasoning processes. The mean time that students spent on the instrument was 13.8 minutes. The protocols were coded by one of the authors using the software program ATLAS.ti (Muhr, 1991), and two expert secondary school history teachers, both of whom held a master's degree in history, reviewed the coding. Coding categories were based on our theoretical framework, and we formulated four primary categories with subcategories, as displayed in Table 1.

Students' Scores on the HPT Instrument
In Table 2, we present the 170 students' mean prior knowledge score, mean category scores (POP, ROA, and CONT), and mean HPT score. We consider mean HPT scores < 2.50 as denoting inadequate ability, scores ≥ 2.50 < 3.00 as denoting adequate ability, scores ≥ 3.00 < 3.50 as denoting good ability, and scores ≥ 3.50 as denoting excellent ability. The students' individual mean HPT scores ranged from 1.56 to 3.89 on a 4-point scale, with a mean score of 3.21. The mean prior knowledge score was 2.23 on a 4-point Notes. POP = present-oriented perspective, ROA = role of the historical agent, CONT = historical contextualization. The mean POP score was calculated using the opposite scoring system of the ROA and CONT categories. The maximum score of 4 for the POP category shows a very low level of presentism. scale. The four prior knowledge questions are presented in the Appendix. The best overall student performance was observed on question one, which asked about the First World War, with 91% of the students answering correctly. The second question, which asked about the rise of Hitler, was answered correctly by 55% of the students, whereas the third question, which asked about the Wall Street Crisis, was answered correctly by only 26% of the students. The last question, which asked about the Treaty of Versailles, was answered correctly by 48% of the students. Using Tukey's honestly significant difference (HSD) post hoc test, we found no significant differences between the average HPT performances of female and male students. Next, we calculated a mean HPT score for students from the same school and used Tukey's HSD post hoc test to determine whether some schools outperformed other schools or scored exceptionally low compared to other schools. The test displayed no significant differences among schools. To examine the possible correlation between students' mean prior knowledge scores and their mean HPT scores, we calculated a Pearson correlation coefficient and found a small but statistically significant correlation of 0.19 at the 0.05 level. In contrast to Hartmann and Hasselhorn (2008), we did not find a significant correlation between students' mean HPT scores and their history grades. Table 3 breaks down the sample by student HPT ability. A mean HPT score ≥ 3.50 was achieved by 32 students (19%), which indicated excellent ability to perform HPT, while only seven students (4%) obtained a mean HPT score < 2.50, which indicated inadequate ability to perform HPT. Most students (n = 81, 48%) achieved mean HPT scores ≥ 3.00 < 3.50, which indicated good ability to perform HPT.

Student Reasoning on the HPT Instrument
To explore how the 15-to 16-year-old pre-university students arrived at their answers, we asked 36 students to think aloud as they solved the HPT instrument. Table 4 displays the individual mean HPT scores combined with students' reported use of the various components of HPT. The highest mean HPT score achieved in this sample was 3.89 on a 4.0 scale (David, Eva, and    Note. Numbers display the total number of references. John), and the lowest was 1.89 (Bas). The mean HPT score for this sample was 3.39, and only two students (6%) received mean HPT scores < 2.50 (Sophie and Bas), while 17 students (47%) achieved mean HPT scores ≥ 3.50.
Viewing the past from a present-oriented perspective. Only one student (Ben) in the sample viewed Hannes as stupid or ignorant. Moreover, this perspective was only apparent when Ben thought about the second item of the HPT instrument ("He will see that only in a democracy can people take part in decision making. He will decide wisely and not choose NSDAP."). Ben applied his historical economic knowledge (high rate of unemployment) but did not include in his reasoning that democracy was uncommon in Germany in the 1930s, which caused Ben to perceive Hannes as being naïve: Honestly, I think that he is too naïve to understand that only in a democracy can people take part in decision making. He is only afraid that his business, more specifically, his fathers' business, is going to be bankrupt. Nobody had a job, and he only wants economic welfare. (Ben, reasoning about item 2) Five other students (Sean, Rose, Peter, Mark, and Sophie) also exhibited a present-oriented perspective when working on the HPT instrument. Though they did not view Hannes as stupid or ignorant, none of them included in their reasoning that the knowledge we have now was not available to people in the 1930s. For example, Rose appeared unaware that Hannes could not have known the outcome (e.g., the beginning of the Second World War) of Hitler's political rise: Rose: He will definitely not vote for the NSDAP. No one can approve of what this party has done to the world. Hitler was responsible for the Second World War.
Interviewer: Is Hannes going to consider this? Rose: Yes, I think so. Hannes might vote for the NSDAP because he is not satisfied with the current government, but I think that he will not vote for the NSDAP because Hitler murdered thousands of people. (Rose, reasoning about item 1) In addition, Sophie, Mark, Sean, and Peter indicated in their reasoning that Hannes knew that voting for the NSDAP would result in violence and terror. Therefore, these students concluded that Hannes could not vote for the NSDAP. Out of the six students who displayed a present-oriented perspective, two students (Rose and Sean) recorded mean HPT scores > 3.50, three students (Ben, Peter, and Mark) achieved mean HPT scores ≥ 3.00 < 3.50, and one student (Sophie) had a mean HPT score < 2.50.
However, most of the students (n = 30) were aware of their possible present-oriented perspective when attempting to explain Hannes' actions. Many students applied their chronological knowledge to emphasize that the information we have now was not available to Hannes at that time. For example, Ryan noted that the scenario was set before the start of the Second World War, and thus, Hannes could not have known the consequences of Hitler's rise to power: The source states that the scenario is set in 1930. Hitler became the political leader of Germany in 1933? I do think so. Hannes is living in 1930, and Hitler became the leader a few years later so he could vote for the NSDAP in 1930, right? Because he does not know what Hitler has done to the world. (Ryan, reasoning about item 1) Another example of the awareness of a present-oriented perspective through the use of chronological knowledge was detected when students reasoned about item 3 ("He will not vote for the NSDAP as their ideas are highly transparent. It is clear that this party wants war.") and concluded that we now know the outcome of the political rise of Hitler, but that people in the past did not have access to this knowledge in the 1930s: I do not think that he knows that the NSDAP might want a Second World War because this scenario is set in 1930. I think that in 1930 he easily could not know that the NSDAP wanted a war. He really could not know it. (Paul, reasoning about item 3) Historical empathy. Of the sampled students, 22 students made 49 affective connections with the historical agent (Hannes) as they explained his actions. In their reasoning, these students included arguments based on recognizable situations and emotions. They seemed to interpret or translate the historical situation into a situation that they could experience today. For example, Mark attempted to explain Hannes' decision by describing a more contemporary situation that he himself could experience as he reasoned about item 4 ("As a member of a wealthy family, he would like to return to the German Empire as his family was better off. Therefore, he will vote for an anti-democratic party."): With respect to item 6, "Because his father's business is almost bankrupt, he might vote for a party that protects small business owners," many students used affective connections to explain why Hannes might vote for the NSDAP. For example, Paul imagined that he himself had financial troubles and thus considered what he would do in a similar situation: I think that this is legitimate. I think that he will vote for the NSDAP. He is going to consider . . . I have a feeling that looking at Hannes' situation, the most important goal for him is that the family business is going well. Looking at myself, I would be happy if my business was making a profit, so I think that this could be the case for Hannes, too. (Paul, reasoning about item 6) Three students (Stella, Tom, and Mark) made five or more affective connections in their reasoning. These students often tried to personalize the historical situation when deciding whether Hannes would vote for the NSDAP. When reasoning about different items, these students used phrases such as "if I were in his shoes" and "I would decide what the best option is for me." Furthermore, 13 of the 49 affective connections (27%) were made by students with mean HPT scores ≥ 3.50. Most affective connections (n = 34, 69%), however, were made by students with mean HPT scores ≥ 3.00 < 3.50, and two affective connections (4%) were made by students with mean HPT scores < 3.00 (see Table 4). Interestingly, though some students indicated that they did not know the specific historical context of Germany, they nonetheless succeeded in answering items correctly by making affective connections. For example, Stella used affective connections to reason that Hannes might long for the period of the German Empire: I think most Germans were better off during the German Empire period, but I do not know the specific circumstances of that period, and if this might have resulted in better economic conditions for his family. Personally, I can understand that you might want the German Empire back because your personal wealth might be higher, and I personally can imagine that is a good thing for everybody. (Stella, reasoning about item 4) Only four students (Jim, Ashley, Ben, and Kim) referenced Hannes' position or his family's position in society in their reasoning. For example, Jim reasoned that Hannes' family was wealthy and respected, and therefore, he might have voted for the NSDAP if this party could ensure the prestige of Hannes' family. However, no student reasoned that Hannes' family might be part of the bourgeoisie and, thus, might long for the German Empire period  when most of these families had far greater political influence.
In addition to making affective connections and considering the role of the historical agent, the protocols revealed yet another type of historical empathy. In particular, one student (Stella) used her knowledge of current values and beliefs of different places in the world. Stella reasoned that in other parts of the world that had a one-party political system, there could be economic welfare and people could be satisfied: Yeah, but maybe he does not feel the need to take part in political decision making. Why would you take part in decision making if you think the government makes wise and good decisions? Decisions that are also good for you. The only thing I then could say every 4 years when there are elections: You are doing a great job, keep up the good work. You still see this in parts of Asia, where people think that they do not need political influence because it is going very well within their own country. (Stella, reasoning about item 2) Reconstructing the historical context. The protocols further revealed that the 36 students used different types of knowledge to reconstruct the historical context as they engaged in HPT. Most references were made to social-political knowledge, with a total of 279 references, followed by 183 references to socialeconomic knowledge, and 154 references to chronological knowledge. Far fewer references were made to social-cultural knowledge (n = 81) and spatial knowledge (n = 5). See Table 4 for detailed information.
The protocols revealed that 32 students made references to chronological knowledge. Four students (Emma, Mark, Sophie, and Bas) did not display any form of adequate chronological knowledge. Three of these students obtained mean HPT scores ≤ 3.00, while one student (Emma) obtained a mean HPT score of 3.33. Two of these students (Mark and Sophie) also viewed the past from a present-oriented perspective. Only five students made references to spatial knowledge. Of these five, one student had a mean HPT score of 3.78 (Nina), three students (Tom, Emma, and Ben) had mean HPT scores of 3.33, and one student (Anna) had a mean HPT score of 3.00. These students, for example, referred to the geographical size of Germany during the period of the German Empire.
Social-economic knowledge was referenced by 35 students in their reasoning when taking a historical perspective. Kim and Anna, for example, both noted that the economic circumstances in Germany in the 1930s resulted in people being attracted to a strong leader who promised increased economic welfare by creating jobs: There [in Germany] was much unemployment, and it is all very bad. The country was first doing alright, but at that time, the economic circumstances were poor and people were dissatisfied with this. Because Hitler was a strong leader and the NSDAP would stimulate the economy by creating jobs, Hannes might vote for the NSDAP. (Kim, reasoning about item 1) I think that he could vote for the NSDAP because he says that he is desperate. He is close to being unemployed, and I think that the NSDAP tried to create more jobs and they will provide a job for Hannes. They [the NSDAP] promised more jobs, and that might result in Hannes voting for the NSDAP. (Anna, reasoning about item 1) All students in the sample displayed social-political knowledge in their reasoning. For example, Kevin stressed the political statements of the NSDAP to explain why Hannes might vote for such a political party: Before the Second World War, the NSDAP was known as a very good political party. Hitler promised many things, and many people believed him because they wanted a better future. I do not know if Hitler was already against the Jews, but he promised a lot more jobs. I think that was very clever, and it resulted in many votes for the NSDAP. (Kevin, reasoning about item 8) Most students in the sample (n = 28) also displayed social-cultural knowledge in their reasoning. These students referenced their knowledge about German cultural behaviors and beliefs in the 1930s. For example, Ashley noted that Hannes might have been influenced by the propaganda spread by Hitler and his political party: Hannes could not see what was really going on in Germany because the German people were getting a very subjective image due to all the Nazi media and propaganda. Therefore, he could not see that the NSDAP wanted a war. (Ashley, reasoning about item 3) Other students referred to the unfair Treaty of Versailles, which resulted in anger among Germans toward a democratic government, while still others mentioned that Hannes was not used to living in a democracy with more freedom but poor economic circumstances. Therefore, he might be skeptical about this type of government.
The two students (Sophie and Bas) with the lowest mean HPT scores (2.11 and 1.89, respectively) demonstrated far less knowledge than did students with higher mean HPT scores. Together with Mark (HPT score of 3.00), they were the only students who combined just two different types of knowledge, specifically, social-economic and social-political knowledge. They did not use chronological, spatial, or social-cultural knowledge. All other students combined at least three types of knowledge in their reasoning. For example, Sean and Stella combined chronological knowledge (e.g., Treaty of Versailles in 1919), social-economic knowledge (e.g., poor economic circumstances), social-political knowledge (e.g., foreign policy of the Nazis), and social-cultural knowledge (e.g., the Germans' anger regarding the Treaty of Versailles) when completing the assessment.
We also calculated a mean score for the use of different knowledge components by totaling the number of references to knowledge and then dividing this sum by five (the number of different knowledge components). For example, Bas made six references to knowledge and obtained a mean score of 1.20, whereas Tom made 38 references to knowledge and obtained a mean score of 7.60. When dividing our sample by the mean HPT score of 3.21 (based on the 170 students' mean HPT score), students with a mean HPT score above 3.21 had an average of 4.26 references to knowledge in the protocols, whereas students with mean HPT scores less than 3.21 made, on average, just 2.34 references to knowledge.
Task approaches. Most of the students (n = 32) explicitly referenced the text about Hannes when working on the instrument, as displayed in Table 4. These students re-read parts of the text or referenced specific information when reasoning about individual items. Only four students read the text once, did not look at it again, and did not explicitly refer to it in their reasoning (see Table 4). Furthermore, most of the students (n = 25) displayed their specification of ignorance, i.e., they doubted their conclusions or indicated that they did not possess the knowledge. These students, for example, were not familiar with the specific political viewpoints of the Nazi Party and did not know how Hannes would react or respond to the instrument's items. Consequently, they had to speculate: Lauren: He will not vote for the NSDAP. Their ideas are easy to see through. It is clear that this party wants a war, but I do not know if Hannes could see this.
Interviewer: Why not?
Lauren: I do not know the ideas of the NSDAP. Was it obvious that Hitler wanted to start a war? I do not know this. (Lauren, reasoning about item 3) Three students, Kim, Rachel, and Anna, explicitly stated in their reasoning that they could not identify the answer because the source did not provide the specific information. One student, Tom, explicitly stated how he was going to approach the task without any encouragement from the interviewer: First, I always examine the assignment before looking at the source. What do I have to do? I see a fill-in assignment with statements that I have to score. Next, I am going to look at the source. Okay, we have a source about Germany in the 1930s. (Tom, before beginning the assessment)

CONCLUSIONS AND DISCUSSION
In this mixed-method study, we tested 170 pre-university students' abilities to perform HPT as evidenced by their performance on an HPT instrument developed by Hartmann and Hasselhorn (2008), and we explored the underlying contextualization processes of 36 pre-university students. In the late 1990s, Angvik and Von Borries (1997) conducted a cross-national survey that aimed to examine 15-to 16-year-old students' views on history education in Europe. One of the questions asked that students place themselves in the position of a young man or woman living in the 15th century who was being forced into marriage. The students were asked what they would do in such a circumstance if they had lived during that time period. Most students participating in the study found it difficult to reconstruct, accept, and acknowledge the concept of a forced marriage, and thus, they often expressed a present-oriented perspective. Nonetheless, only seven of the 170 students (4%) participating in our study had a mean HPT score < 2.50 out of a maximum 4.00 score, indicating inadequate ability to perform HPT. Most students (n = 81, 48%) achieved a mean HPT score ≥ 3.00 < 3.50, indicating good ability to perform HPT. This finding is consistent with those of Huijgen et al. (2014) and Hartmann and Hasselhorn (2008), who also examined 15-to 16-year-old pre-university students' abilities to perform HPT.
Furthermore, our analysis of the verbal protocols of 36 students indicated that five students viewed the past from a present-oriented perspective in that they did not realize that people in the past did not have the same knowledge that we have today. One student viewed Hannes as naïve and reasoned at the lowest level of the Lee and Ashby (2001) taxonomy such that people of the past are regarded as ignorant or stupid. In contrast, the other 30 students were aware of the consequences of their present-oriented perspective when explaining historical agents' decisions, a finding consistent with that of Berti et al. (2009).
Various studies and handbooks on teaching and learning history emphasize that presentism restricts historical understanding and that many students might view the past from a present-oriented perspective (e.g., Haydn, Stephen, Arthur, & Hunt, 2015;Lévesque, 2008;Seixas & Morton, 2013). However, the causes of presentism exhibited by students are typically not described. Based on our findings, we posit that historical knowledge plays a critical role in preventing presentism. Scholars, such as Brooks (2013), VanSledright (2001), Van Boxtel and Van Drie (2012), and Wineburg (2001), have suggested a relationship between historical content knowledge and students' ability to perform HPT, and our study seems to confirm this association. Specifically, we found a small but significant correlation (0.19) between students' prior chronological knowledge and their performance on the HPT instrument. Furthermore, students' protocols indicate that students who displayed good or excellent ability to perform HPT (mean HPT score ≥ 3.00) used more historical topic knowledge, particularly chronological and social-political knowledge but also social-cultural and social-economic knowledge, in their reasoning than did students with mean HPT scores < 3.00. Compared to the lowest-performing students (Sophie and Bas), students who demonstrated good and excellent abilities to perform HPT also employed more types of knowledge in their reasoning.
Making affective connections with a historical agent (e.g., if the students' own fathers had money problems) could also facilitate individuals as they engaged in HPT (Endacott & Sturtz, 2014;Virja & Kouki, 2014), and our data seem to confirm this. However, five of the six students who displayed a present-oriented perspective also made affective connections. This suggests that making affective connections alone might not prevent presentism but that, to prevent such presentism, affective connections must include the role of the historical agent and the broader historical context. Based on the protocols, we also contend that students are skipping specific scaffolds that include the role of the historical agent when working on HPT tasks, as few students explicitly considered the role of the historical agent (e.g., Hannes' wealthy family influenced his preference for conservative political parties). Teachers who focus their instruction on teaching students to combine affective connections while also considering the role of the historical agent and the broader historical context might find that their students' performance on HPT tasks improves rather significantly.
Consistent with a relevant point raised by Berti et al. (2009), we found evidence that students used the affective element of historical empathy when they did not succeed in reconstructing the historical context. Some students explicitly noted that they did not know the specific historical circumstances but could understand Hannes' decision to vote for the Nazi Party because they, too, would not want to be unemployed. Because we observed this in the reasoning of only a few students, historical empathy as a fallback rationale and the interaction between affective and cognitive processes of historical empathy when performing HPT require further research. Future research could also focus on whether affective connections are more difficult to make when the historical topics or issues are more distant, such as dating back to ancient Rome or the Middle Ages. Again, further research is needed to examine the extent to which students can perform HPT by evaluating current beliefs and values of different parts of the world, as we found one student in our study who applied this strategy.
Another finding of our study regards the instrument itself. While testing students about their knowledge of historical facts is rather easy, valid and reliable instruments that measure students' historical reasoning competencies are scarce. Consequently, scholars have argued for new assessment formats (Ercikan & Seixas, 2015;Reich, 2009), and the development of the History Assessments of Thinking (HATs) is a good example (Breakstone, 2014;Breakstone, Smith, & Wineburg, 2013). That said, we used an instrument validated by Hartmann and Hasselhorn (2008) and Huijgen et al. (2014). However, some limitations with regard to the instrument's validity and practical improvements must be noted. First, students may have misread or misunderstood two of the instruments' items. We noted in the protocols that some students explicitly struggled with answering the instruments' first and sixth items. For example, when working on the first item, students had to check the first box ("Does not fit his situation at all") to receive the maximum score. However, these students reasoned that Hannes could not have known the outcome of the Second World War, thus displaying a non-present-oriented perspective, and therefore ticked the last box ("Fits his situation very well"), which yielded a score of 1. The same thing occurred when students answered the sixth item. Rewriting these two items might reduce the potential for misunderstanding, and more detailed instruction on the terminology of the scoring boxes (such as the inclusion of a test item) before beginning the instrument might resolve this problem. Second, we observed in the data protocols that students answered the instruments' items after closely reading and investigating the source. Testing students' reading comprehension levels could better identify the impact student reading ability has on students' scores on the HPT instrument.
Furthermore, the ROA items on the instrument require examination. Originally, Hartmann and Hasselhorn (2008) conceptualized the ROA items as an intermediate stage between presentism and historical contextualization. Students could refer to roles or institutions that they know from their own lives (e.g., the role of a father or businessman). However, we did not find evidence that the ROA items represented an intermediate stage between presentism and historical contextualization. To further investigate the relationships between the cognitive and affective elements of historical empathy and the instrument's ROA items, the ROA items could be divided into two categories, specifically, items that might trigger more affective processes of historical empathy (e.g., "If my own father would be fired, I could vote for the Nazi Party in the 1930s.") and items that might trigger more cognitive processes (e.g., "Hannes belongs to a wealthy family. Therefore, he could vote for the Nazi Party.").
Though we did not find a strong relationship between generic task approaches (e.g., evaluating decisions, expressing doubt) and domain-specific strategies, such as performing HPT, quasi-experimental studies that focus on promoting HPT and include generic task approaches could provide valuable insights for the teaching and learning of history. Furthermore, more quasiexperimental research involving the spatial context and the position of a historical agent in society is needed as only a few students in our study displayed this in their reasoning. Thus, it would be interesting to see whether teacher instruction focused explicitly on the spatial context and the historical agents' position results in better HPT performance.
One study limitation is that we conducted exploratory research that included only 36 students in our thinking aloud protocols and focused on only one historical topic with one related assignment. The instrument's scenario was also fictional. Thus, more research on how students perform HPT when addressing real historical sources or other tasks about agents' decisions is needed. Furthermore, in the quantitative portion of our study, we included only four questions about students' chronological knowledge to measure prior knowledge. As this is a further limitation, future research should focus on the relationship between one's ability to perform HPT and one's prior knowledge and should include more questions on different types of historical knowledge to confirm the relationship we found when analyzing the thinking-aloud protocols. Another limitation is that our mean HPT score for the qualitative sample was slightly higher (3.39) than the mean HPT score for the quantitative sample (3.21). Furthermore, as we only included pre-university students in our study, it would be interesting to compare their HPT ability with students' HPT abilities at other educational levels and to examine possible differences in students' specific needs to successfully perform HPT.
Finally, we discuss some practical implications for the teaching of history. Although the majority of the students in our study did not view the past from a present-oriented perspective, six students did do so. To decrease students' presentism, Huijgen and Holthuis (2015) presented a sample lesson about the rise of Hitler that was shaped by the theory of constructive controversy (Johnson & Johnson, 2009). Teachers could use these types of exercises to evaluate students' presentism and to scaffold historical contextualization. Furthermore, our study, similar to those of Levstik (2011) and Havekes et al. (2012), indicated that building extensive and different frames of reference could help students perform HPT. However, as Reisman and Wineburg (2008) noted, it does not result in the automatic application of historical knowledge, as HPT also requires a deep understanding of the difference between past and present. Accordingly, this is not an easy task for teachers, as Levstik and Groth (2002) noted. However, lessons combining historical contextualization with historical empathy tasks could promote this understanding. Recently, Endacott and Pelekanos (2015) presented a good example of such a lesson when teaching a unit on ancient Athens. As shown in this study, HPT is a complex process, but structural attention and classroom practice can promote students' understanding of the past and help them prepare to participate in a civic society.

NOTE
1 Other terms to describe the concept of historical perspective taking (HPT) include perspective recognition (e.g., Barton & Levstik, 2004;Brooks, 2011) and rational understanding (e.g., Lee & Ashby, 2001). Although we acknowledge the difficulty of taking a perspective of a historical agent, we chose to use the term HPT to match the terminology of previous studies using the HPT instrument (Hartmann & Hasselhorn, 2008;Huijgen et al., 2014).