Various sociological methodologies are used when designing and executing research. Each of these methods, including comparative-historical sociology, ethnometho-dology, ethnography, evaluation research, qualitative methods, and survey research, has strengths and weaknesses. While debate surrounds qualitative versus quantitative methods, the best sociological research often integrates both kinds of methods to test hypotheses.
Most nineteenth-century social scientists, including Emile Durkheim, Herbert Spencer, and Karl Marx, engaged in analyses of historical data and made cross-cultural comparisons in their studies of human society. The work of these early historical sociologists was guided by the belief that societies were evolving and that the western European societies were the most advanced. The premise was that societies progressed via evolution and that progress was good. Comparisons were used as a tool for the development of social facts based on cross-cultural and/or historical data. In modern times cross-cultural comparisons serve to provide a better understanding of the structures and institutions of different societies.
The primary strength of comparative-historical research is its use of an interdisciplinary approach. If the scope conditions are clear and the criteria are specified and defined, then this approach is an important method for obtaining "social facts."
Data available for cross-cultural and historical analyses face multiple hurdles. To illustrate, one must remember that information from a culture is embedded in the language, status sets, and expectations for the use of the data, as well as the time and place where the data were collected. There is always the issue of making sure that data sets are comparable and that the variables are equivalent. One primary limitation noted by Etienne Van de Walle (2005) is that although historical demographers have access to volumes of information, they are frequently limited by only including information on elite male populations with little or no information about females or the common man.
QUALITATIVE METHODS AND ETHNOGRAPHY
The primary qualitative methods sociologists use are ethnography interviews and direct observations. Interviews with research participants may range from open-ended interviews with flexible content directed by the interviewer to more structured questions asked by multiple researchers; in the latter case there is an obvious requirement for internal consistency so that all interviewers ask the same questions in the same way, and hopefully obtain comparable data. Researchers engaged in direct observations may have varying levels of participation, ranging from covert observation to participant observation where the researcher becomes an active member of the group.
Many ethnographers agree that to fully understand a complex social situation, one must enter into an unbiased observation or interaction with the society being studied. William Foote Whyte (1955) argued in his classic Street Corner Society: The Social Structure of an Italian Slum, that the only way to describe a society is to live in it, learn to speak the language, and participate in its social events and everyday life. Some of the most well known ethnographies have been guided by similar principles, for example, Elliott Liebow’s Tally’s Corner (1967), Margery Wolf’s The House of Lim (1968), and Laud Humphrey’s Tearoom Trade (1975), to mention only a few. A quantitative counting using preconceived survey questions only provides answers to the questions and could well be biased by the selection process as well as by the perceived social desirability of the responses by the researcher. In contrast, a qualitative analysis provides detailed description and information and new perspectives necessary for hypothesis development. An issue that should be addressed is the role of the researcher in the ethnographic research and whether or not an external observer can really study the internal workings of a society without bias. Also, a weakness of ethnographic field studies is generalization. But this weakness is often resolved by integrating the qualitative results of the fieldwork with quantitative results obtained in research in which a large population is systematically and randomly sampled and surveyed (see the work of Knodel, Chamratrithirong, and Debavalya  as an example of such integration).
The term ethnomethodology was first used in the 1960s by Harold Garfinkel (1967) in research determining how people make sense of their worlds. Garfinkel noted that for interactions to be smooth, everyday communication and interpersonal interactions have to be based on prior assumptions. Ethnomethodologists commonly study the normal through the use of techniques such as conversation analysis and breaching experiments, which force an examination of the usual, accepted, and unquestioned. The documented reactions of others to these experiments confirm which behaviors are normative (Cohen 2006).
The strength of ethnomethodology is that it permits the researcher to analyze the normal. For example, Allen Smith and Sherry Kleinman (1989) use narratives to demonstrate the patterns of discourse in conversations, which can be used to train medical personnel in the delivery of bad news and desexualizing gynecological exams. The weakness is that assumptions about what is normal and what is expected are in continual flux so that general-izability is sometimes limited.
Organizational sociologists, following a long-standing pos-itivistic agenda, often use evaluation research to determine whether the programs and routines of such groups as corporate organizations, social agencies, and educational institutions actually perform as planned. Evaluation research techniques involve formative research; setting the agenda, goals, and strategies for the organization; determining how these can be quantified and hence evaluated; and summa-tive evaluation, determining if these quantifiable outcomes of both the steps and the goals meet the predetermined standards. Evaluation researchers usually use multiple techniques, including ethnography and survey instruments (see Rossi, Lipsey, and Freeman 2003). The strength of evaluation research is that it is used to minimize expenses while improving the quality of the accepted standards set by formative research. One weakness is that organizations have multiple systems and the research may not target the critical part of the systems. Organizations are in continual flux so that their evaluations must be ongoing and easily modifiable to respond to changing conditions.
Survey research involves the "systematic gathering of information on a defined social group" (Rapley and Hansen 2006, p. 616). The group is typically sampled from a larger population; information is obtained by asking standard questions about previously operationalized variables. One reason for the use of survey research is its simplicity; if one wants to know information, ask. Questions may be either closed-ended or open-ended. The strengths of the analysis and the generalizations from the findings are determined in part by the sample size and selection. Samples may range from small convenience samples to large randomized representative samples. Surveys may be administered via interviews, mailed questionnaires, telephone calls or online. Don Dillman (2000) argues that mail surveys using the "tailored design method," a detailed methodology of multiple contacts ensuring compliance, often have the greatest likelihood of being understood, completed, and returned; these are all characteristics necessary for the survey results to be truly representative of the population.
Surveys are usually used after one has developed hypotheses to be tested quantitatively. The best-known surveys have an efficient methodology and obtain accurate and current information about the population. Examples include the U.S. Current Population Survey, the World Fertility Surveys, and the U.S. National Surveys of Family Growth.
A strength of survey research, if done correctly, is its potential for strong and generalizable statistical analysis. However, the best surveys require randomization, adequate sample size, and a high completion rate. It must be remembered, however, that survey methods provide only a "partial description of complex social issues____They are but one tool, of many, in the [social scientist's] armamentarium" (Rapley and Hansen 2006, p. 617).