Unit 3 Basic designs in sociology

1. What are the basic designs in sociological research?

The three basic designs in sociological research are surveys, experiments, and ethnography. Surveys involve systematically gathering answers to standardised questions from a specified sample of respondents. Respondents may be asked to answer questions by mail, over the phone, or in face-to-face interviews. Surveys are particularly useful for gathering information about events that cannot be measured directly.

One must clearly understand the sample and population employed in a survey. The population is the total number of people who share a characteristic that the sociologist wishes to study. For example, students at both public and private colleges and universities may constitute a population. A sample is a limited but representative subset of a population. For example, we might take a random sample of all students at private and public colleges and universities in the United States at a particular point in time. In a random sample, everyone within the population has an equal chance of being selected. A random sample is far more important than a large sample; small random samples almost always give more accurate results than large but nonrandom samples. In assessing published research, one must determine the nature of the sample employed. Frequently magazine surveys involve large numbers of respondents, but the samples tend to be nonrandom and include only those who subscribe to or purchase the magazine.

The wording and sequence of the questions in a survey may affect the validity and reliability of the data obtained. Using different words to measure the same concept may sometimes yield different results. The order in which questions are asked may also affect the responses. Issues raised in earlier questions can affect how respondents think about later questions. The form of the response is also relevant. In a closed response question respondents must choose from the set of answers provided by the researcher. In an open response question respondents answer in their own words. Using both forms for the same topic may produce different results.

Questionnaires are efficient for studying a large number of geographically scattered people. For some types of research, however, sociologists need more information than a short questionnaire provides; they turn to interviews in such situations. Interviews can be conducted in person or by phone. Effective interviewers listen carefully, record the responses accurately, know when to probe for more information, and know that the validity and reliability of interviews depend on the interac­tion between the interviewer and the respondent. Some interviews are semistructured: the general and specific issues to be covered are identified in advance, but the respondents are free to talk about each topic in terms most meaningful to them. Structured interviews are those in which the wording and sequence of questions are carefully planned in advance. In an unstructured interview, the questions and the topic are not predetermined; the interviewer and the subject engage in free-flowing conversation.

The experiment offers the most effective technique for establishing cause-and-effect relation­ships. Experiments, particularly laboratory experiments, provide greater control over the indepen­dent and dependent variables and over the grouping of subjects into experimental and control groups. The independent variable is assumed to be the causal factor and is systematically manipu­lated. The dependent variable is the factor affected by the manipulation of the independent vari­able. Laboratory experiments generally involve an artificial situation that can be regulated carefully by the researcher. That is, many factors can be held constant while the potential independent variable is isolated.

Laboratory experiments sometimes raise ethical questions. For example, researchers must frequently deceive subjects regarding the purpose of the experiment. Critics have raised questions about the right to deceive people or to cause them anxiety or humiliation in the name of science. Most colleges and universities have established clear guidelines on ethical principles. Researchers must 1) explain the experiment to the subject, 2) minimise lying, 3) warn subjects about potential hazards, 4) describe the use of the data, 5) maintain confidentiality, and 6) inform subjects fully before seeking consent.

Field experiments occur in real-world settings. Manipulating potential independent variables is more difficult in these experiments because the controlled conditions of the laboratory are absent. As a result, sociologists sometimes use natural social settings and naturally occurring phenomena to answer research questions. Occasionally they are able to separate subjects into experimental and control groups; sometimes such groups are established after the fact. Such limitations, however, impose restrictions on the extent to which the researcher can make causal assertions. Ethical issues also emerge in field experiments. For example, experimental groups operating with the independent variable may benefit in a way that the control groups do not. Deception, privacy, and other ethical issues also enter the picture.

In ethnographies, sociologists do not attempt to influence or change people's behaviour. They simply observe people in everyday settings, usually over a long period, and provide detailed descrip­tions and interpretations of social life as it happens. In overt participant observation, the researcher participates in the social life of those who are being studied, and his or her role is made known to those people. Occasionally, however, such knowledge may affect the subjects' behaviour. Covert participant observation remedies this problem because sociologists do not identify themselves as researchers; instead they try to act like members of the group they are studying. This method, however, raises serious ethical questions. Is it ethical to deceive people deliberately by pretending to belong to their group when the true intent is to study them? Privacy issues also emerge. A variation on covert participant observation is nonparticipant observation, in which sociologists do not participate in the activities of the group under study. They simply observe the group in its everyday setting.

2. What other approaches do sociologists employ?

Historical studies enable sociologists to examine the changes that take place in institutions, groups, and societies over time. Such studies are particularly appropriate in studying phenomena that occur infrequently. They are also useful in studying events that unfold over a long period. Finally, histori­cal research enables sociologists to draw on documents created for other purposes. Secondary analysis is the analysis of data that were originally collected for another reason. It helps shed new light on historical events.

Sociologists also employ content analysis to uncover information in both historical and contem­porary materials. Content analysis involves various types of recorded communication, such as letters, diaries, laws, novels, and newspapers. It is especially useful in historical studies because it provides a way to organise and summarise systematically both the manifest and the latent content of the material. Content analysis may involve both qualitative research (research that depends on inter­pretations by the researcher) and quantitative research (research that employs statistics).