Social science involves the study of people. Social research methods are based on a systematic approach to studying social phenomena. Social sciences include anthropology, sociology, economics, psychology, political science, communications, and history. Researchers in each of these disciplines are interested in pursuing explanations for human social behavior. Sociologists and anthropologists, in particular, are interested in the empirical study of human behavior from a cultural and societal perspective. Oftentimes, social scientists are critiqued for their lack of a vigorous set of research standards that tend to be more apparent in the natural science disciplines. However, social scientists adopt both quantitative and qualitative methodologies by following a measurement process that helps to ensure both reliability and validity in their work. In addition, researchers develop studies that focus on exploratory, descriptive, and explanatory societal questions.
Research Design and Measurement in Social Research
Dimensions of research design include exploration, description, and explanation. Each design supports different fundamental research goals. Exploratory research is used when the goal is to become familiar with basic facts, people, and concerns. Researchers are often able to develop a well-grounded mental picture of the social situation that is unfolding. Researchers also generate main ideas and develop tentative theories, determine the feasibility of doing additional research, formulate questions and issues for further inquiry, and develop techniques and a sense of direction for research.
Descriptive research is used to provide an accurate profile of a group as well as describe processes, mechanisms, or relationships. In addition, descriptive research gives a verbal or numerical picture of a social setting. Researchers use the descriptive approach to find information to stimulate new explanations; present basic background information or context; create a set of categories and classify types; and clarify a sequence, set of stages, or steps.
Researchers carry out explanatory research when they are interested in solving a problem or establishing an explanation for why something exists as it does. A further goal of explanatory research is to determine the accuracy of a principle or theory. Explanatory research can also be used to find out which competing explanation is better and advance knowledge about an underlying process or social relationship. Explanatory research is especially useful to build or elaborate on a theory so it is more complete, extend a theory or principle into new areas or issues, and provide evidence to support or refute an explanation.
In addition to considering the type of design in social research, researchers must be able to develop measurement standards to establish consistency and accuracy in their concepts. Measurement involves developing a set of clearly defined variables that can be compared and analyzed as precise data points. In order to develop clear, well-defined concepts, researchers must conceptualize and operationalize their variables. Conceptualization involves developing a precise definition of the idea being studied. This definition serves as the standard for empirical measurement in the social world. Without this step, it would be difficult to develop a standard for the concepts being observed. The standard helps to ensure that the measure is observed consistently and accurately throughout the research study.
A second step in the measurement process is operationalization. Researchers operationalize concepts by specifically outlining how the measure will be tested empirically. In this process, it is imperative that researchers outline data collection steps. How do we capture the concept empirically? Whether we plan to test our concepts using surveys, participant observation, in-depth interviews, archival research, or experiments, we need to clearly establish guidelines for collecting data.
Measurement varies among quantitative and qualitative research designs. The process outlined above tends to be more relevant to quantitative research designs, where the research process is more deductive and linear. In quantitative designs, the researcher usually develops a hypothesis and then seeks empirical observations that either confirm or refute the hypothesis. In contrast, qualitative research designs usually depend on a grounded theory model, where researchers use an inductive approach to developing theoretical explanations. In this case, researchers may develop a research plan that requires they gather empirical evidence prior to developing theoretical statements or hypotheses. These designs tend to be nonlinear and more flexible in how they are carried out.
Developing a strong research study requires a strong research proposal. Effort spent in this process helps the researcher outline a research agenda that combines both a discussion of relevant past research and proposed contributions to the field. A proposal should include the following sections: a clear and well-defined research question or problem, a discussion of relevant literature and past research studies, a description of how data will be collected including a discussion of how variables have been conceptualized and operationalized, and a conclusion that includes ethical issues and points for further discussion.
The proposal creates a map of sorts for the researcher to follow. With a well-defined plan of study, the researcher can enter his or her study process with a clear idea of the overall agenda.
Specific Research Designs
Traditionally, social science researchers work with both quantitative and qualitative frameworks. Over time, emphasis has been placed on one research design or the other. However, more recently, there has been a push for collaboration among research designs. Different levels of depth, detail, and understanding can be accomplished through a combination of designs. For example, a field researcher concentrating on participant observation can collect rich, detailed descriptions of what she sees in the field, but may not have the data in a format that allows for any quantification for comparison. By including a survey tool, she can then compare basic demographic characteristics or other data that suits her research needs. Triangulation of methods has become accepted practice. Most social scientists agree that a combination of methods assists in helping to ensure validity and reliability.
Because of the nature of the data that is collected, most research methods are inherently quantitative or qualitative. Quantitative methods include survey research, secondary analysis, existing statistical analysis, and experiments. Qualitative methods include field research, historical comparative research, and natural experiments. Focus groups, case studies, and content analysis can be used as either quantitative or qualitative depending on how they are employed and the suitability of the research question.
Consistently increasing in the use since the 1930s and the development of George Gallup’s Gallup Poll, survey research is one of the most popular means of collecting data in the social sciences. If they are directed to a representative target population, surveys have shown to be useful in measuring popular opinion. Surveys are also used to measure attitude, behavior, knowledge, and consumer preference. Popular tools across the social sciences, surveys provide large amounts of data that can be gathered quickly, are relatively inexpensive to collect, and are easily quantified to be used for statistical analysis.
However, there are drawbacks to this method as well. Because most surveys consist of close-ended questions, the depth and breadth of the response is limited. Respondents may be choosing the ‘best’ response, though it may not truly represent their feelings or ideas. In addition, survey analysts have to account for a standard margin of error that can be derived from analyzing cases of missing responses, deceit, or simply misunderstanding the nature of the question.
Two additional quantitative methods that are often used in conjunction are secondary analysis and existing statistics. Each method relies on past collection of data. Secondary analysis is used when a researcher is interested in re-analyzing a study that has already been carried out. This occurs to offer a test of reliability of the study but can also serve as a means to test other cause-and-effect relationships or theoretical explanations of the variables.
Using existing statistics is slightly different. In this method, researchers are not collecting the data first hand either, but instead develop their own analytic schemata for data that has been compiled by another source. For example, researchers use U.S. Census data to analyze educational attainment by neighborhood by using zip code aggregate data and educational attainment rates by high school.
The experimental method is used most often in psychology but has been used to isolate causal factors in both sociology and anthropology. Used to isolate factors of cause-and-effect relationships, experiments are highly detailed in their design and rely on the researcher’s ability to test for validity through control of the experiment conditions. Most experiments in the social sciences rely on a controlled, simulated environment, where participants are observed before and after a change to the independent variable.
Qualitative methods differ in that the form of data is not numeric but highly detailed and relies on an incorporation of thematic analysis and empirical observation. Examples of data include, but are not limited to, interview transcriptions, maps, photographs, diaries or journals, oral histories, and document analysis. Common qualitative methods include field research, historical comparative research, and natural experiments.
Field research is based on techniques of observation and analysis in a natural setting. The goal of field research is to capture a real picture of human social behavior. In order to do this, researchers study groups and individuals in their natural social settings. Researchers can take on numerous roles in the field. Roles are categorized by four distinctions: participant observer, complete participant, complete observer, and observer participant. As participant observer, the researcher acts as a pseudomember of the group. Members are aware of the researcher’s role and recognize the duality of their responsibility as participant and observer. When the researcher acts as a complete participant she acts as a member of the group, keeping her identity as a researcher hidden from the group. There is some question as to what level of deception is acceptable. As a complete observer, the researcher remains undetected and unnoticed, reporting observations. The fourth role the researcher assumes is the observer as participant. In this situation, the researcher is a known, overt observer, however she has limited contact with members. She remains slightly detached from the group, which some argue allows for greater objectivity.
Two issues emerge in terms of validity of field research data. Researchers need to consider internal consistency, or whether the data are plausible given all that is known about a person, group, or event, not considering forms of human deception. External consistency, or whether observations and data can be verified or cross-checked with other sources of data, needs to be considered as well. Both of these issues rest with the research design and operationalization of the research question.
Historical comparative research is another form of qualitative research. Sources for this data include archives, library collections, and memory studies. Historical comparative researchers look for three things in their evidence: implicit conceptual framework, particular details, and empirical generalizations. The conceptual framework should include an implicit understanding of assumptions and perspective of the person reporting events. Researchers need to take careful consideration of times, dates, events, chronology, people involved, and locations. It is essential that researchers try to verify information by focusing on factual statements on which there is agreement. However, consideration must be taken to ensure that data is valid and not simply the result of a hidden historical agenda. The researcher is also interested in interpretations of evidence and the “silences” or cases where the evidence fails to address a particular event, topic, or issue.
Researchers need to be aware of their own cultural biases. Without this awareness, judgments of other cultures may be made that jeopardize the validity of the data. Even though it is difficult at times, one way to do this is to defocus before entering the field. This step involves figuratively removing oneself from the primary focus of the research paradigm. The goal is to remain open to everything one may encounter and to not take occurrences for granted but to acknowledge everything as potentially important and meaningful.
Social research methods allow us to provide explanations of human social behavior. Through a systematic approach, researchers can begin to make sense of the social world. Although some disciplines focus on particular methods over others (psychology: experiments; sociology: surveys; anthropology: field research), each takes into account the goals of the discipline in an effort to collect the most reliable and valid results possible.
- Babbie, E. (2002). The basics of social research (2nd ed.). Belmont, CA: Wadsworth.
- Creswell, J. (2002). Research design: Qualitative, quantitative, and mixed method approaches. Thousand Oaks, CA: Sage.
- Neuman, L. (2003). Social research methods: Quantitative and qualitative approaches (5th ed.). Boston: Allyn & Bacon.
- Wysocki, D. K. (2004). Readings in social research methods. Belmont, CA: Wadsworth.