You are here

Building students' research expertise: history as a case study

Rosemary Hipkins

This article, the third in a series about carrying out research as a student learning activity, discusses how research relates to context—in this case, history.

The challenge is to help students experience and understand research as the basis for new knowledge construction in any discipline area.

Journal issue: 

Building students’ research expertise: history as a case study

Rosemary Hipkins


This is the third in a series of articles about carrying out research as a student learning activity. The first article reported data, drawn from the New Zealand Council for Educationsl Research’s Learning Curves project, on secondary students’ experiences of carrying out research projects that would be assessed for the National Certificate of Educational Achievement. In the course of this study we found that it was common for students to see research as a process of information retrieval and repackaging. Many of them felt they had not been taught the skills they needed to carry out such projects (Hipkins, 2005a). That first article suggested that when teachers also view research as a set of information retrieval and repackaging skills they may not see the need to teach and assess research skills in their subject.

The second article gave a short overview of “information literacy” and suggested that information literacy skills need to be developed in every subject, not just in one or two. Some aspects of information literacy require specific conceptual and contextual knowledge in a subject area, in addition to more generic information skills and necessary dispositions such as open-mindedness and criticality. The second article raised questions such as: On what basis would students evaluate the trustworthiness of a source if they knew nothing about the discipline area in question? (Hipkins, 2005b).

This article takes the discussion one step further—how research relates to context—by identifying research skills that are quite specifically related to a particular type of knowledge system, the discipline of the historian. But it is not just for history teachers; I have chosen historical research simply as a convenient example. In the previous articles I argued that we need to move beyond seeing research as a process of finding out and reporting back (no matter how critically the information is evaluated). The challenge is to help students experience and understand research as the basis for new knowledge construction in any discipline area.

Gilbert (2001) describes three important levels of literacy. The first, functional literacy, is fairly obvious— being able to read and write printed text. This is the foundation on which the other levels of literacy are built. The second level is cultural literacy, but this term does not have the everyday meaning of the word “cultural”; it means being able to see a knowledge system (history, for example) as a culture in itself. Gilbert describes this as the ability to use and understand “the structures and systems through which words achieve their meaning” (p. 180). She goes on to say:

Literacy, at this level, is, therefore, context-specific: one cannot just be literate, one has to be literate with respect to something—some aspect of knowledge in a particular culture. Texts can only be interpreted if something is known about the context—the ‘rules of the game’ or the ‘discourse’—within which they were produced, and within which they were intended to be understood.

(Gilbert 2001, p. 180)

Gilbert says that individuals achieve the third level— critical literacy—when they can actively participate in the knowledge-building processes of the discipline and transform knowledge within it. One way to help students learn the “rules of the game” of knowledge construction is to give them opportunities to do carefully supported research projects that provide a feel for the way specialists in a discipline work. In this way they can begin to build cultural literacy, and perhaps even move on to true critical literacy in time. But this is not research as a process of information retrieval. Nor can it just draw on generic information literacy skills. As the above quotation makes clear, this type of literacy is discipline-specific, and needs to be built afresh in every important discipline area.

In what follows I use research in history as an example of the types of teaching and learning challenges entailed in learning a discipline at a “meta” level—that is, as “the rules of the game” of a knowledge system.

Learning from research about doing research in history

I need to state at the outset that I am not, and never have been, a history teacher. I make no claims about how the competencies I discuss should be taught (or whether they are currently taught at all in schools). Rather, I am using the example of history to introduce the more general case that research requires discipline-specific competencies. If students are to learn about research, then they will need opportunities to practise and discuss these competencies.

I chose history because the literature search we carried out when preparing the first article turned up some very interesting pieces of educational research about the challenges of learning how to do research in this discipline. I find it interesting that other disciplines yielded very few articles that discussed research about learning to do research at school. To make it easier to keep track of the general points of the argument, I begin by briefly sketching three key research articles on which I drew. All are American in origin.

Learning to use multiple sources of information

In this project, a team of American researchers (Stahl, Hynd, Britton, McNish & Bosquet, 1996) worked with 44 l0th-grade students. Some students were given an assignment that required them to research and report factually on an event from the Vietnam War. The other students researched the same event, using the same sources of information, but were asked to state an opinion on their findings. The students were given free choice from a range of carefully chosen documents. These included short, succinct historical overviews, biographies, records of key meetings, and so on. The researchers investigated which sources the students used, and in what ways. As we shall see, their findings led them to make what might seem to be counter-intuitive recommendations about the relationship between students’ research activities and their learning of new content.

Comparing research skills in history and science

In a small but revealing case study, one researcher (Fehn, 1997) worked with 16 American science students who had been identified at the national level as gifted and talented at carrying out science investigations. The students were observed as they explored and talked about five documents that were carefully chosen to cover a range of source types and perspectives on slavery in the American South. The students then orally told the researcher a narrative about slavery, drawing from whichever of these five sources they chose. Fehn analysed the stories to seek instances where the students used the competencies needed to create historical narratives from source documents (these competencies are described below). He found that being talented at scientific investigation does not confer the skills needed to carry out good research in history. Just two of the 16 students demonstrated some of the necessary competencies, and they had both completed an advanced Baccalaureate course in history.

Comparing school students and historians

One method of investigating expertise in a particular discipline is to compare the performance of experts and amateurs. In this project, “think aloud” responses were collected from eight university historians and eight academically talented senior secondary students. They were each given the same task— to make sense of one small episode from the American Civil War. To do this they had access to a series of historical documents that included paintings, diaries, and textbook accounts (Wineburg, 1991). As in both the research projects described above, these documents were carefully chosen to give conflicting perspectives and some differences of detail. Wineberg’s findings were very similar to those reported for both of the other two projects. He identified a set of research competencies that the historians used but the school students did not. While it would obviously be unrealistic to expect school students to behave like expert historians, the differences found by Wineberg add to the picture of possible targets for actively teaching discipline-specific research skills.

Learning to be a history researcher: what could be taught?

These three research projects give consistent messages about four sets of competencies that seem to be specific to history (and probably to related sub-disciplines such as biography writing). I now describe each of these in turn.

Selecting sources of information

All three research projects described a competency called a “sourcing heuristic”. Wineberg summarised it like this: “When evaluating historical documents, look first at the source or attribution of the document.” (Wineburg, 1991, p. 79, emphasis added).

In a warm-up activity when working with the science students, Fehn modelled a process of checking source details that were provided on the back of each history resource. But when it came to carrying out their own interpretive exercise, the science students mostly ignored this information when they selected the sources they would use to build their narrative. They simply forgot to do it.

Wineberg commented that the school students in his project often checked sources after they had explored an item, whereas the experienced historians did this before anything else. This gave them important information with which to “read” the text and weigh up the meaning of the details they discovered there. He commented that the students missed important subtleties in each text1 because they spent a lot of the reading time trying to place the document in context—which checking the source could have established. He also commented that:

Students seemed to view texts as vehicles for carrying information in which the attribution was just the last thing to be read, one more bit of information to be added to the other bits that had been gathered. But for historians who used attributions to erect elaborate scenarios about authors and circumstances of document generation, the attribution was not another bit of information but the ‘bit’ from which all else emanated. Historians seemed to view texts not as vehicles but as people, not as bits of information to be gathered but as social exchanges to be understood.

(Wineburg, 1991, p. 83)

A third of the students in Stahl et al.’s (1996) project chose as their first source a book about the Vietnam War, written from a partisan point of view by a retired army colonel; some of these students looked no further. Both Fehn and Stahl reported student naiveté about the unquestioned reliability of some sources, especially textbooks, and Wineberg commented at some length on this. One of the historians in his project described the shortcomings of textbook accounts as (among others):

•&&being short on detail

•&&having a tendency to be patriotic and hence political;

•&&giving information of the sort that could be used to answer multiple-choice questions; and

•&&blurring over things that are not clear.

It seems this brevity and (false) clarity has an appeal for students that may inhibit the development of their history research skills. Stahl’s team found that, given the free choice of a range of different types of documents, many of their students went for historical overviews that were short, succinct, and made clear assertions. The students took detailed notes from the first source they had chosen. Thereafter they appeared to learn little more content, adding to notes mostly when these corroborated what had already been gleaned. They tended to ignore conflicting evidence and were not good at winnowing relevant information from longer, less coherent authentic sources (such as transcripts of actual meetings).

To what extent and how students can be taught this sourcing competency, I leave it for history teachers to say. My point is that it is discipline-specific, and there are definite aspects to learn and practice.

Contextualising the sources found

Contextualisation—situating text in the temporal and spatial context of its time and place—is an important aspect of interpreting historical texts. Wineberg (1991) summarises this competency as: “When trying to reconstruct historical events, pay close attention to when they happened and where they took place” (p. 80).

This requires the history researcher to look closely at documents to find their layers of meaning. Wineberg says trained historians focus on:

What precedes and follows events, on how long they lasted, and on the amount of time between their occurrence and their recording by witnesses. … The ‘where’ of this heuristic is concerned with situating events in concrete spaces and determining the conditions of their occurrence—issues of geography, weather, climate and landscape.

(Wineburg, 1991, p. 80)

For experienced historians, some of this contextual detail will begin to build once they have sourced the document, simply because of other knowledge they already hold. This existing knowledge allows them to create hypotheses about what they will find as they explore further. It would be unrealistic to expect that students could carry out this kind of sophisticated analysis without help. But by now, it should come as no surprise that the school students in these studies did little of this type of intellectual work at all. Indeed, commenting on their handling of the three different historical paintings they had to hand, Wineberg observed that the students seemed to see these as some sort of multiple-choice option, with the challenge being to pick the one they preferred. The historians, by contrast, attended carefully to all the differences the contextual clues provided, trying to gain a better sense of the messy, sometimes conflicting stories about the same event.

Corroborating information from different sources

The use of multiple sources also potentially allows for corroboration by comparing and contrasting sources. Again, Wineberg (1991) provides us with a succinct sense of this competency: “Whenever possible, check important details against each other before accepting them as plausible or likely” (p. 77).

Fehn combines this competency with contextualisation (see above) and calls it “interpretive acumen”. He says it is about attitudes and values. It involves open-mindedness and a willingness to use more than one source, remaining alert to different points of view and differing interpretations in different sources. For example, one of Fehn’s resources was a copy of a street poster that announced penalties for slave-owners who shirked their share of night patrols in the area. This presented the possibility of two interesting counter-narratives. It suggested that not all slave-owners were willing suppressers of their slaves’ liberty, and that not all slaves accepted the suppression of their liberty passively. Fehn’s science students did not “see” this type of evidence, as an experienced historian would be expected to do, and so they largely missed the differences of perspective in the materials they were given. Most of them went straight for the longest, most formal written source and based their account on that alone.

Some students in the Learning Curves focus groups (see Hipkins, 2005a) had interesting perspectives on the requirement to use multiple sources in research projects. They essentially saw this as busy work—something they had to do to demonstrate that they had not been “lazy”, even if they had already found the information they wanted within one main source. Here again we see a difference between information retrieval and repackaging and the challenge of practising the actual knowledge-building competencies used by historians. Pointing to another teaching challenge, Fehn found that the students thought written sources were reliable but visual sources were likely to give a more biased interpretation of events. As he pointed out, all text, whether written or visual, is an interpretation of some sort. Again, Wineberg succinctly takes up the same point:

The question put by the historian to the source was not ‘Is the source biased?’ But ‘How does a source’s bias influence the quality of its report?’ Students, on the other hand, seemed to view bias as a binary, an attribute of some texts but not of others.

(Wineburg, 1991, p. 84

Making generalisations

Part of the skill of the historian lies in juxtaposing documents to generate themes that can then be woven together to build a complicated narrative. The use of multiple sources as evidence to support such themes or generalisations is an important indicator of quality in the research. Reworking evidence into a multilayered but clear story seems to demand yet another set of sophisticated competencies. To what extent can students begin to learn these at school? One consideration seems to be the way the research assignment is framed.

Stahl’s team found that those students who were asked to write a description stuck close to the order of information presented in the first text they read. But students asked to write an opinion tended to draw on prior knowledge and to ignore what they had learned from any of the sources provided. These researchers concluded that writing persuasive essays that use evidence to support assertions is a skill that needs to be taught. First, of course, the assignment must be framed in a way that does not prompt simple information retrieval and repackaging. And that brings us to the question of whether students should be expected to use research as a way of learning content that is new for them.

Subject knowledge and literacy

In her article on the three different types of literacy, Gilbert notes that many people interpret the idea of “scientific literacy” (for example) as being about knowing the big ideas of science—that is, about having a good overview of content knowledge in the discipline (Gilbert, 2001). I hope it is clear from the case studies above that this is not what literacy means in the cultural or critical sense. Knowing at a “meta” or systems level involves learning at least as much about the processes of knowledge building as about the knowledge itself. But that does not mean that gaining knowledge is not important. It just needs to be thought about in different terms.

On one level it is obvious that having at least some background knowledge of the historical events to be researched is a helpful start for shaping an inquiry framework for a history project. (I note in passing that experts who are building new knowledge are seen as expert precisely because they already hold deep knowledge of the area in which they are researching.) Four of Wineburg’s historians were experts in American history, but four were not. Even though these others did not have the same content knowledge, their use of the competencies outlined above gave them a set of tools to bring to the task so that they were able to compensate for this lack of background (and, indeed, were aware that they needed to do so).

So how much should students know about a topic before they begin their research? I have already noted that Stahl’s team found that when students “research” to learn content they tend to add little that is new if their first source is informative and easy to understand. In this situation they also ignore conflicting evidence. For this reason, these researchers recommended that students gain at least an overview of the relevant content knowledge in the area before beginning the research process. From the perspective of research as information retrieval and repackaging, this might seem like a counter-intuitive idea. Learn the topic first and then do the research? However, if the research competencies are as complex as this article has suggested, and if fostering them is seen as the main aim of the research, then the suggestion makes sense.

The history teachers whose educational research I have introduced here carefully selected the sources to be used by the students with whom they worked. They didn’t leave the students to find sources for themselves, but they did expect them to select for themselves, and to justify their selection. In this way, the teachers’ greater expertise was absolutely necessary in selecting material that provided opportunities for students to demonstrate (or not) their history research competencies.

Developing a researcher identity

If the main purpose of doing research, as outlined in the introduction, is to build new types of literacies, teachers do need to rethink the types of learning outcomes they want students to achieve. Gilbert’s article makes the extent of the challenge clear. She says students need “a well-developed sense of self or an ‘identity’ with respect to this context: that is, some sense of themselves as a ‘thinker’, or a ‘scientist’, or ‘investigator’, or to at least be able to imagine this as a possibility” (Gilbert, 2001, p. 187, emphasis in the original). We could add, in the context chosen here, the need to gain a sense of themselves as “historians”. The only way to build that is by doing history research in the same way that historians do, albeit with teacher support as outlined above.

This article is the third in a series that responds to the common argument that research need not be assessed in other subjects if students learn how to do it in English. I hope it is clear by now that I strongly disagree with this position. If students are to understand how history is actually interpreted by historians, they need to learn to interpret it, too—or at least to be able to imagine the processes that would be followed in doing so. If they are to understand how scientific theories are created, they need to learn that process as well. And, as we have seen above, being an able science student does not automatically confer the combination of knowledge, skills and values needed to be a good history researcher. Research competencies are discipline-specific, and they have to be learned in context.


Fehn, B. (1997). Historical thinking ability among talented math and science students: An exploratory study. Paper presented at the Annual Meeting of the American Research Association, Chicago, March 28.

Gilbert, J. (2001). It’s science, Jim, but not as we know it: Re-thinking an ‘Old’ discipline for the ‘Knowledge Society’. SAMEpapers, 174–190.

Hipkins, R. (2005a). Students’ experiences of “researching” in different subjects. set: Research Information for Teachers, 1, 18–22.

Hipkins, R. (2005b). Information literacy and student research. set: Research Information for Teachers, 2, 27–31.

Stahl, S., Hynd, C., Britton, B., McNish, M., & Bosquet, B. (1996). What happens when students read multiple source documents in history? Reading Research Quarterly, 31(4), 430–456.

Wineburg, S. (1991). Historical problem solving: A study of the cognitive processes used in the evaluation of documentary and pictorial evidence. Journal of Educational Psychology, 83(1), 73–87.


1&&&Note that “text” here means material in the widest sense—i.e., paintings and other visual documents, not just written accounts. Hence “reading” means exploring the meaning of the details found, not just a more literal decoding of words.

Rosemary Hipkins is a senior researcher at the New Zealand Council for Educational Research.