Mouly (1978) states that while historical research cannot meet some of the tests of the scientific method interpreted in the specific sense of its use in the physical sciences (it cannot depend, for instance, on direct observation or experimentation, but must make use of reports that cannot be repeated), it qualifies as a scientific endeavour from the standpoint of its subscription to the same principles and the same general scholarship that characterize all scientific research.
Historical research has been defined as the systematic and objective location, evaluation and synthesis of evidence in order to establish facts and draw conclusions about past events (Borg, 1963). It is an act of reconstruction undertaken in a spirit of critical inquiry designed to achieve a faithful representation of a previous age. In seeking data from the personal experiences and observations of others, from documents and records, researchers often have to contend with inadequate information so that their reconstructions tend to be sketches rather than portraits. Indeed, the difficulty of obtaining adequate data makes historical research one of the most taxing kinds of inquiry to conduct satisfactorily.1 Reconstruction implies an holistic perspective in that the method of inquiry characterizing historical research attempts to ‘encompass and then explain the whole realm of man’s [sic] past in a perspective that greatly accents his social, cultural, economic, and intellectual development’ (Hill and Kerber 1967).
Ultimately, historical research is concerned with a broad view of the conditions and not necessarily the specifics which bring them about, although such a synthesis is rarely achieved without intense debate or controversy, especially on matters of detail. The act of historical research involves the identification and limitation of a problem or an area of study; sometimes the formulation of an hypothesis (or set of questions); the collection, organization, verification, validation, analysis and selection of data; testing the hypothesis (or answering the questions) where appropriate; and writing a research report. This sequence leads to a new understanding of the past and its relevance to the present and future.
The values of historical research have been categorized by Hill and Kerber (1967) as follows:
- It enables solutions to contemporary problems to be sought in the past.
- It throws light on present and future trends.
- It stresses the relative importance and the effects of the various interactions that are to be found within all cultures.
- It allows for the revaluation of data in relation to selected hypotheses, theories and generalizations that are presently held about the past.
The particular value of historical research in the field of education is unquestioned. Although one of the most difficult areas in which to undertake research, the outcomes of inquiry into this domain can bring great benefit to educationalists and the community at large. It can, for example, yield insights into some educational problems that could not be achieved by any other means. Further, the historical study of an educational idea or institution can do much to help us understand how our present educational system has come about; and this kind of understanding can in turn help to establish a sound basis for further progress of change. Historical research in education can also show how and why educational theories and practices developed. It enables educationalists to use former practices to evaluate newer, emerging ones. Recurrent trends can be more easily identified and assessed from an historical standpoint – witness, for example, the various guises in which progressivism in education has appeared. And it can contribute to a fuller understanding of the relationship between politics and education, between school and society, between local and central government, and between teacher and pupil.
Choice of subject
As with other methods we consider in this book, historical research may be structured by a flexible sequence of stages, beginning with the selection and evaluation of a problem or area of study. Then follows the definition of the problem in more precise terms, the selection of suitable sources of data, collection, classification and processing of the data, and finally, the evaluation and synthesis of the data into a balanced and objective account of the subject under investigation. There are, however, some important differences between the method of historical research and other research methods used in education. The principal difference has been highlighted by Borg (1963), who suggests that in historical research, it is important for the student to define carefully the problem and appraise its appropriateness before moving into earnest into the project, as many problems may not be suitable for historical research methods, while, on the other hand, other problems may have little or no chance of yielding any significant results either because of the dearth of relevant data or because the problem is trivial.
One can see from Borg’s observations that the choice of a problem can sometimes be a daunting business for the potential researcher. Once a topic has been selected, however, and its potential and significance for historical research evaluated, the next stage is to define it more precisely, or, perhaps more pertinently, delimit it so that a more potent analysis will result. Too broad or too vague a statement can result in the final report lacking direction or impact. Best (1970) expresses it like this: ‘The experienced historian realizes that research must be a penetrating analysis of a limited problem, rather than the superficial examination of a broad area. The weapon of research is the rifle not the shotgun’. Various prescriptions exist for helping to define historical topics. Gottschalk (1951) recommends that four questions should be asked in identifying a topic:
- Where do the events take place?
- Who are the people involved?
- When do the events occur?
- What kinds of human activity are involved?
As Travers (1969) suggests, the scope of a topic can be modified by adjusting the focus of any one of the four categories; the geographical area involved can be increased or decreased; more or fewer people can be included in the topic; the time span involved can be increased or decreased; and the human activity category can be broadened or narrowed. It sometimes happens that a piece of historical research can only begin with a rough idea of what the topic involves; and that delimitation of it can take place only after the pertinent material has been assembled.
Hill and Kerber (1967) have pointed out that the evaluation and formulation of a problem associated with historical research often involve the personality of the researcher to a greater extent than do other basic types of research. They suggest that personal factors of the investigator such as interest, motivation, historical curiosity, and educational background for the interpretation of historical facts tend to influence the selection of the problem to a great extent.
One of the principal differences between historical research and other forms of research is that historical research must deal with data that already exist. Hockett (1955) argues that, as history is not a science which uses direct observation as in chemistry or biology, the historian, like the archaeologist, has to interpret past events by the traces which have been left. Of course, the historian has to base judgements on evidence, weighing, evaluating and judging the truth of the evidence of others’ observations until the hypothesis explains all the relevant evidence.
Sources of data in historical research may be classified into two main groups: primary sources, which are the life-blood of historical research, and secondary sources, which may be used in the absence of, or to supplement, primary data.
Primary sources of data have been described as those items that are original to the problem under study and may be thought of as being in two categories. First, the remains or relics of a given period: although such remains and artefacts as skeletons, fossils, weapons, tools, utensils, buildings, pictures, furniture, coins and objets d’art were not meant to transmit information to subsequent eras, nevertheless they may be useful sources providing sound evidence about the past. Second, those items that have had a direct physical relationship with the events being reconstructed: this category would include not only the written and oral testimony provided by actual participants in, or witnesses of, an event, but also the participants themselves. Documents considered as primary sources include manuscripts, charters, laws, archives of official minutes or records, files, letters, memoranda, memoirs, biography, official publications, wills, newspapers and magazines, maps, diagrams, catalogues, films, paintings, inscriptions, recordings, transcriptions, log books and research reports. All these are, intentionally or unintentionally, capable of transmitting a firsthand account of an event and are therefore considered as sources of primary data. Historical research in education draws chiefly on the kind of sources identified in this second category.
Secondary sources are those that do not bear a direct physical relationship to the event being studied. They are made up of data that cannot be described as original. A secondary source would thus be one in which the person describing the event was not actually present but who obtained descriptions from another person or source. These may or may not have been primary sources. Other instances of secondary sources used in historical research include: quoted material, textbooks, encyclopedias, other reproductions of material or information, prints of paintings or replicas of art objects. Best (1970) points out that secondary sources of data are usually of limited worth because of the errors that result when information is passed on from one person to another.
For a detailed consideration of the specific problems of documentary research, the reader is referred to the articles by Platt (1981) where she considers those of authenticity, availability of documents, sampling problems, inference and interpretation.
Because workers in the field of historical research gather much of their data and information from records and documents, these must be carefully evaluated so as to attest their worth for the purposes of the particular study. Evaluation of historical data and information is often referred to as historical criticism and the reliable data yielded by the process are known as historical evidence. Historical evidence has thus been described as that body of validated facts and information which can be accepted as trustworthy, as a valid basis for the testing and interpretation of hypotheses. Historical criticism is usually undertaken in two stages: first, the authenticity of the source is appraised; and second, the accuracy or worth of the data is evaluated. The two processes are known as external and internal criticism respectively, and since they each present problems of evaluation they merit further inspection.
External criticism is concerned with establishing the authenticity or genuineness of data. It is therefore aimed at the document (or other source) itself rather than the statements it contains; with analytic forms of the data rather than the interpretation or meaning of them in relation to the study. It therefore sets out to uncover frauds, forgeries, hoaxes, inventions or distortions. To this end, the tasks of establishing the age or authorship of a document may involve tests of factors such as signatures, handwriting, script, type, style, spelling and place-names. Further, was the knowledge it purports to transmit available at the time and is it consistent with what is known about the author or period from another source? Increasingly sophisticated analyses of physical factors can also yield clues establishing authenticity or otherwise: physical and chemical tests of ink, paper, parchment, cloth and other materials, for example. Investigations in the field of educational history are less likely to encounter deliberate forgeries than in, say, political or social history, though it is possible to find that official documents, correspondence and autobiographies have been ‘ghosted’, that is, prepared by a person other than the alleged author or signer.
Many documents in the history of education tend to be neutral in character, though it is possible that some may be in error because of these kinds of observer characteristics. A particular problem arising from the questions posed by Travers (1969) is that of bias. This can be particularly acute where life histories are being studied. The chief concern here, as Plummer (1983) reminds us, resides in examining possible sources of bias which prevent researchers from finding out what is wanted and using techniques to minimize the possible sources of bias.
Researchers generally recognize three sources of bias: those arising from the subject being interviewed, those arising from themselves as researchers and those arising from the subject–researcher interaction (Travers 1969).
Writing the research report
The writing of the final report is equally demanding and calls for creativity and high standards of objective and systematic analysis. Best (1970) has listed the kinds of problems occurring in the various types of historical research projects submitted by students. These include:
1. Defining the problem too broadly.
2. The tendency to use easy-to-find secondary sources of data rather than sufficient primary sources, which are harder to locate but usually more trustworthy.
3. Inadequate historical criticism of data, due to failure to establish authenticity of sources and trustworthiness of data. For example, there is often a tendency to accept a statement as necessarily true when several observers agree. It is possible that one may have influenced the others, or that all were influenced by the same inaccurate source of information.
4. Poor logical analysis resulting from:
- oversimplification – failure to recognize the fact that causes of events are more often multiple and complex than single and simple
- overgeneralization on the basis of insufficient evidence, and false reasoning by analogy, basing conclusions upon superficial similarities of situations
- failure to interpret words and expression in the light of their accepted meaning in an earlier period
- failure to distinguish between significant facts in a situation and those that are irrelevant or unimportant.
5. Expression of personal bias, as revealed by statements lifted out of context for purposes of persuasion, assuming too generous or uncritical an attitude towards a person or idea (or being too unfriendly or critical), excessive admiration for the past (sometimes known as the ‘old oaken bucket’ delusion), or an equally unrealistic admiration for the new or contemporary, assuming that all change represents progress.
6. Poor reporting in a style that is dull and colourless, too flowery or flippant, too persuasive or of the ‘soap-box’ type, or lacking in proper usage.
To conclude on a more positive note, Mouly (1978) itemizes five basic criteria for evaluating historical research:
1. Problem: Has the problem been clearly defined? It is difficult enough to conduct historical research adequately without adding to the confusion by starting out with a nebulous problem. Is the problem capable of solution? Is it within the competence of the investigator?
2. Data: Are data of a primary nature available in sufficient completeness to provide a solution, or has there been an overdependence on secondary or unverifiable sources?
3. Analysis: Has the dependability of the data been adequately established? Has the relevance of the data been adequately explored?
4. Interpretation: Does the author display adequate mastery of his data and insight into the relative significance? Does he display adequate historical perspective? Does he maintain his objectivity or does he allow personal bias to distort the evidence? Are his hypotheses plausible? Have they been adequately tested? Does he take a sufficiently broad view of the total situation? Does he see the relationship between his data and other ‘historical facts’?
5. Presentation: Does the style of writing attract as well as inform? Does the report make a contribution on the basis of newly discovered data or new interpretation, or is it simply ‘uninspired hack-work’? Does it reflect scholarliness?
The use of quantitative methods
By far the greater part of research in historical studies is qualitative in nature. This is so because the proper subject-matter of historical research consists to a great extent of verbal and other symbolic material emanating from a society’s or a culture’s past. The basic skills required of the researcher to analyse this kind of qualitative or symbolic material involve collecting, classifying, ordering, synthesizing, evaluating and interpreting. At the basis of all these acts lies sound personal judgement. In the comparatively recent past, however, attempts have been made to apply the quantitative methods of the scientist to the solution of historical problems (Travers 1969). Of these methods, the one having greatest relevance to historical research is that of content analysis, the basic goal of which is to take a verbal, non-quantitative document and transform it into quantitative data (Bailey 1978).
Content analysis itself has been defined as a multipurpose research method developed specifically for investigating a broad spectrum of problems in which the content of communication serves as a basis of inference, from word counts to categorization. Approaches to content analysis are careful to identify appropriate categories and units of analysis, both of which will reflect the nature of the document being analysed and the purpose of the research. Categories are normally determined after initial inspection of the document and will cover the main areas of content.
The purposes of content analysis have been identified by Holsti (1968):
1. to describe trends in communication content
2. to relate known characteristics of sources to messages they produce
3. to audit communication content against standards
4. to analyse techniques of persuasion
5. to analyse style
6. to relate known attributes of the audience tomessages produced for them
7. to describe patterns of communication.
The life history, according to Plummer (1983), is frequently a full-length book about one person’s life in his or her own words. Often, Plummer observes, it is gathered over a number of years, the researcher providing gentle guidance to the subject, encouraging him or her either to write down episodes of life or to tape-record them. And often as not, these materials will be backed up with intensive observations of the subject’s life, with interviews of the subject’s friends and acquaintances and with close scrutiny of relevant documents such as letters, diaries and photographs. Essentially, the life history is an ‘interactive and co-operative technique directly involving the researcher’ (Plummer 1983).
Miller (1999) demonstrates that biographical research is a distinctive way of conceptualizing social activity. He provides outlines of the three main approaches to analysis, that is to say:
1. the realist, focusing upon grounded-theory techniques
2. the neo-positivist, employing more structured interviews
3. the narrative, using the interplay between interviewer and interviewee to actively construct life histories.
Denzin (1999) suggests that there are several varieties of biographical research methods including:
5. narrative writing
6. personal history
7. oral history
8. case history
9. life history
10. personal experience
11. case study.
This is addressed further by Connelly and Clandinin (1999) who indicate several approaches to narrative inquiry:
1. oral history
3. annals and chronicles
5. memory boxes
This involves the researcher both in selecting an appropriate problem and devising relevant research techniques. Questions to be asked at this stage are first, ‘Who is to be the object of the study?’ – the great person, the common person, the volunteer, the selected, the coerced? Second, ‘What makes a good informant?’ Plummer (1983) draws attention to key factors such as accessibility of place and availability of time, and the awareness of the potential informant of his or her particular cultural milieu. A good informant is able and willing to establish and maintain a close, intimate relationship with the researcher. It is axiomatic that common sympathies and mutual respect are prerequisites for the sustenance and success of a life history project. Third, ‘What needs clarifying in the early stages of the research?’ The motivations of the researcher need to be made explicit to the intended subject. So too, the question of remuneration for the subject’s services should be clarified from the outset. The issue of anonymity must also be addressed, for unlike other research methodologies, life histories reveal intimate details (names, places, events) and provide scant cover from prying eyes. The earlier stages of the project also provide opportunities for discussing with the research subject the precise nature of the life history study, the logistics of interview situations
and modes of data recording.
Central to the success of a life history is the researcher’s ability to use a variety of interview techniques. As the occasion demands, these may range from relatively structured interviews that serve as general guides from the outset of the study, to informal, unstructured interviews reminiscent of non-directive counseling approaches espoused by Carl Rogers (1945) and his followers. In the case of the latter, Plummer (1983) draws attention to the importance of empathy and ‘nonpossessive warmth’ on the part of the interviewerresearcher. A third interviewing strategy involves a judiciousmixture of participant observation and casual chatting, supplemented by note-taking.
Typically, life histories generate enormous amounts of data. Intending researchers must make early decisions about the use of tape-recordings, the how, what and when of their transcription and editing, and the development of coding and filing devices if they are to avoid being totally swamped by the materials created. Readers are referred to Fiedler’s (1978) extensive account of methods appropriate to field studies in natural settings.
Reliability in life history research hinges upon the identification of sources of bias and the application of techniques to reduce them. Bias arises from the informant, the researcher and the interactional encounter itself: Several validity checks are available to intending researchers. Plummer (1983) identifies the following:
1. The subject of the life history may present an autocritique of it, having read the entire product.
2. A comparison may be made with similar written sources by way of identifying points of major divergence or similarity.
3. Acomparisonmaybemadewithofficial records by way of imposing accuracy checks on the life history.
4. A comparison may be made by interviewing other informants.
Essentially, the validity of any life history lies in its ability to represent the informant’s subjective reality, that is to say, his or her definition of the situation. Detailed personal accounts and life histories can be interrogated thematically (e.g. the work of Thomas and Znaniecki 1918). Indeed the use of biographies, autobiographies, fictional accounts or newspaper journalism raises the question of what counts as legitimate research data. Perhaps such accounts may be better used to provide sensitizing concepts and contexts rather than as mainstream research data.
Plummer (1983) provides three points of direction for the researcher intent upon writing a life history. First, have a clear view of who you are writing for and what you wish to accomplish by writing the account. Are you aiming to produce a case history or a case study? Case histories ‘tell a good story for its own sake’ (Plummer 1983). Case studies, by contrast, use personal documents for wider theoretical purposes such as the verification and/or the generation of theory. Second, having established the purpose of the life history, decide how far you should intrude upon your assembled data. Intrusion occurs both through editing and interpreting. Editing (‘cutting’, sequencing, disguising names, places etc.) is almost a sine qua non of any life history study. Paraphrasing Plummer, editing involves getting your subject’s own words, grasping them from the inside and turning them into a structured and coherent statement that uses the subject’s words in places and your own, as researcher, in others, but retains their authentic meaning at all times. Third, as far as the mechanics of writing a life history are concerned, practise writing regularly. Writing, Plummer observes, needs working at, and daily drafting, revising and redrafting is necessary. For an example of life history methodology and research see Evetts (1991).
Documents take amultitude of forms, including, for example:
- field notes
- diaries and journals
- formal records
- technical documents
- minutes of meetings
- samples of students’ work
- memos and emails
- reports and statistics
- pamphlets and advertisements
- prospectuses and directories
- annals and chronicles
- photographs and artifacts
- conversations and speeches
- policy documents
- primary and secondary sources
- newspaper articles
- books and articles
- public records.
Reliability and validity in documentary analysis
Validity may be strong in first person documents or in documents that were written for a specific purpose (Bailey 1994: 317). However, that purpose may not coincide with that of research, thereby undermining its validity for research purposes. We mentioned earlier the problem of bias, selectivity, being written for an audience and purposes different from those of the researcher, attrition and selective survival; all these undermine validity. In historical research great care is paid to authenticity and provenance, and documents may be subject to chemical analysis here (e.g. of inks, paper, parchment and so on) in order to detect forgeries. Bailey (1994: 318) suggests that face validity and construct validity in documents may be stronger and more sufficient than other forms of validity, though corroboration with other documents should be undertaken wherever possible.
With regard to reliability, while subjectivity may feature highly in certain documents, reliability by corroboration may also be pursued. The standards and criteria of reliability have to be declared by the researcher. Scott (1990) suggests four criteria for validity and reliability in using documents: authenticity; credibility (including accuracy, legitimacy and sincerity); representativeness (including availability and which documents have survived the passage of time); and meaning (actual and interpreted).
Cohen, Louis, Lawrence Manion & Keith Morrison. (2007). Research Methods in Education. London & New York: Routledge.