The Methods of Sociological Research
Four research methods are widely employed in sociological investigation. A method is a strategy for carrying out research in a systematic way – comparable to a blueprint used in building or a recipe in cooking. The four methods discussed here are all expressions of the logic of science. No method is in an absolute sense better or worse than any other; each has characteristic strengths and weaknesses so that any method is particularly suited for certain kinds of research.
The logic of science is clearly expressed in the experiment – a method that seeks to specify a cause-and-effect relationship among variables. Experimental research, in other words, is explanatory in character, attempting to show what factors in the social world cause change to occur in other factors. Experiments are typically based on the text of a specific hypothesis – a theoretical statement of a relationship between independent and dependent variables. The goal of an experiment is to find out whether or not the hypothesis is supported by empirical evidence.
A survey is a method of contacting individuals in order to obtain responses to a series of items or questions. Surveys are particularly useful when we are seeking answers to specific questions, especially when what we want to know cannot be observed directly, such as the political preferences and religious beliefs of individuals, patterns of sexual attraction, or the private lives of married couples. Because surveys typically involve the number of different variables, they (like experiments) are appropriate for conducting explanatory research, in which we attempt to specify the relationship among several variables, seeking correlations or even causal links among them.
Selecting the subjects who will be contacted is only the first step in carrying out a survey. Also required is a specific way to ask questions and record answers. Two commonly used techniques are questionnaires and interviews.
A questionnaire is a series of questions or items to which all subjects are asked to respond. In most cases, the respondent is provided with possible responses to each item, so that the process of answering only involves selecting the best response (the format is similar to multiple-choice examination questions). A questionnaire that provides a set of responses to the subject has a closed-ended format.
In some cases, however, a researcher might want to let a subject respond in an entirely free way. In an open-ended format the subjects are able to express their responses however they wish, which allows subtle shades of opinion to come through. Of course, the researcher later has to make sense out of what can be a bewildering array of answers.
Most often, a questionnaire is mailed to respondents who are asked to complete the form and then return it to the researcher, usually also by mail. This technique is called a self-administered survey. When subjects respond to such questionnaires, no researcher is present, of course; so the questionnaire must be prepared in an attractive way, with clear instructions and questions that are easy to understand. In self-administered surveys, it is especially important to pretest the questionnaire with a small group of people before sending it to all subjects in the study. The small investment of time and money involved can help prevent the costly problem of finding out too late – that instructions or questions were not clear to respondents.
Researchers may also use the interview (sometimes called an interview-survey), which is a questionnaire administered personally to the subject by the researcher. Interviews are especially useful if the items have an open-ended format because the researcher can ask follow-up questions, both to probe a bit more deeply and to clarify the subject’s responses. The researcher must be careful not to influence a subject’s responses, however; sometimes even raising an eyebrow as someone begins to answer a question can be enough to change a response. The advantage of an interview is that a subject is more likely to complete a questionnaire in the presence of a researcher. One disadvantage is that tracking people down is often a difficult job, and more than one attempt may be necessary. Another is that if all subjects do not live in the same area, the costs of conducting research in this way can become extremely high.