By no means all method textbooks discuss document analysis because there is disaccord as to whether this is an independent technique or whether it merely aims to apply different methods to a particular investigation material. There is also controversy as to what has to be understood by “document.” Used in a wider sense as a synonym for the word “source” in historical research, this concept encompasses all objects manufactured by man. However, as a rule, only written sources are considered to be documents. Examples are: reports, the wordings of laws, the minutes of proceedings, records, letters, diaries, leaflets, press articles, books, organization charts, or work instructions. Since these documents have their origin in everyday contexts and there is no contact between the researcher and the production of the investigation material, document analysis is a non-reactive method. Consequently, neither the researcher nor the aim of the investigation has any influence on the material itself. In addition, the costs of document analysis are lower than those of any primary survey. Against this, the method has two major disadvantages: usually, documents only account for parts of the questions asked by the researcher, and they leave room for interpretations.
Document analysis is the obvious choice whenever the objects of study (persons, groups, or organizations) are not available – either because they no longer exist (historical aspects) or because they refuse to be questioned or observed (like fringe groups and elites). In addition to that, document analysis can prepare for the use of other methods or be complementary to them. In the case of written material, document analysis allows for the use of all the methods that social science and the humanities have developed for the study of texts. However, one must always bear in mind that documents are never made without a definite intention. Consequently, the context of their production must not be disregarded. Methodological literature names numerous rules for dealing with documents that are evocative of the criticism of sources in historiography. The trustworthiness of documents has always to be questioned, and the relationship between the wording and the facts that are being described should be regarded carefully, by examining information about the document’s origin and its purpose, as well as the organization that produced the document and the intentions of the author. Quality criteria applied in document analysis are: transparency, diversity, and comparability. As far as comparison is concerned, data obtained with the help of other methods can be used in addition to other documents.
This approach will be illustrated by way of opinion polls carried out in the precomputer age. Usually, the results of such surveys are available in written form. Unlike records, diaries, and other documents, opinion polls provide information about behavioral patterns, opinions, and values across the entire population, as well as about the differences between existing social groups and milieus. When interpreting such findings in the context of a document analysis, one has to ask who commissioned the polling, what was the position taken by the opinion research institute, what was the aim of the survey, who were the recipients of the reports, and whether parallel and preceding studies exist. Additionally, it is essential to submit the polling methods used in that time to a critical review. The validity of the data depends on the instruments being used, as well as on the quality of the random sample, on the way the survey is conducted, on the indicators being implemented, and on the quality of the data analysis. Further methodological aspects to be considered are the questionnaire design and the way the questions are formulated, as well as the interviewers’ motivation and the behavior of the respondents. In order to permit classification of the information contained in the historical report, one has to consider other data – for instance, the results of opinion polls held in other fields of inquiry and over other periods of time – or documents that might provide further indications for answering the research question (such as official statistics, diaries, or reports on public opinion).
- Allport, G. W. (1942). The use of personal documents in psychological science. New York: Social Science Council, Bull. No. 49.
- Angell, R. C., & Freedman, R. (1966). The use of documents, records, census materials, and indices. In L. Festinger & D. Katz (eds.), Research methods in the behavioural sciences. New York: Holt, Rinehart and Winston, pp. 300–326.
- Hyman, H. H. (1972). Secondary analysis of sample surveys: Principles, procedures and potentialities. New York: John Wiley.