: Evaluating Bias in Research
davd.c2
doing_sociological.docxDOING SOCIOLOGICAL
RESEARCH
Sociological research requires careful preparation and a
Clear plan that guides the work. The purpose of a sociological research project may be to obtain preliminary knowledge that will help formulate a theory, or to evaluate an already-existing theory about society and social life. As part of the strategy, the researcher selects from a variety of research methods—specific techniques for systematically gathering data. In the following sections, we look at a range of research methods, including their advantages and disadvantages. We also discuss how you might prepare a sociological research project of your own.
SOCIOLOGICAL RESEARCH METHODS
Sociologists employ a variety of methods to learn about the social world (Table 2.4 on page 38). Since each has strengths and weaknesses, a good research strategy may be to use several different methods. If each yields similar findings, researchers are more likely to have confidence in the results. The principal methods are the survey, fieldwork—either participant observation or detached observation—experiments, working with existing information, and participatory research
SURVEY RESEARCH
A survey relies on a questionnaire or interviews with a group of people in person or by telephone or e-mail to determine their characteristics, opinions, and behaviors Surveys are versatile
38
Chapter 2: Discover Sociological Research and sociologists often use them to test a theory or simply to gather data. Some, such as National Opinion Research Center questionnaires, consist of closed-ended questions for which the respondent checks his or her preferred response. Others, such as the University of Chicago’s Social Opportunity Survey, consist of open-ended questions that permit the respondent to answer in his or her own words
An example of survey research conducted for data collection is the largest survey in the nation, the U.S. Census, which is conducted every 10 years. The census is not designed to test any particular theory. Rather, it gathers voluminous data about U.S. residents that researchers, including sociologists, use to test and develop a variety of theories usually, surveys are conducted on a relatively small number of people, a sample, selected to represent a population the whole group of people to be studied. The first step in designing a survey is to identify your population of interest. Imagine doing a study of behavioral factors that affect grades in college: Who would you survey? Members of a certain age group only? People in the airline industry? Pet owners? To conduct a study well, we need to clearly identify the survey population that will most effectively help us answer the research question. In your study you would most likely choose to survey students now in college, because they offer the best opportunity to correlate grades with particular behaviors. Once they have identified a population of interest, researchers will usually select a sample, as they seldom have the time or money to talk to all the members of a given population, especially if it is a large one. Other things being equal, larger samples better represent the population than smaller ones. However, with proper sampling techniques, sociologists can use relatively small (and therefore inexpensive) samples to represent large populations. For instance, a well-chosen sample of 1,000 voters can be used to represent 100,000 U.S. voters with a fair degree of accuracy, enabling surveys to make election predictions with reasonable confidence. Sampling is also used for looking at social phenomena such as drug or alcohol use in a population: CNN reported recently that 17% of high scholars drink, smoke, or use drugs during the school day, based on a 1,000-student sample polled by the National Center on Addiction and Substance Abuse at Columbia University (Azuz, 2012).Ideally, a sample should reflect the composition of the population we are studying. For instance, if you wanted to be able to use research data about college students to generalize about the entire college student population of the United States, you would want to collect proportional samples from two-year colleges, four-year colleges, large universities, community colleges, online schools, and such. It would not be adequate to survey only students at online colleges or only female students at private four-year schools. To avoid bias in surveys, sociologists may use random sampling, whereby everyone in the population of interest has
A Sample: A portion of the larger population selected to represent the whole.
Population: The whole group of people studied in sociological Research.
Random sample:
A sample in which everyone in the population of interest has an equal chance of being chosen for the study.
Equal chance of being chosen for the study. Typically, they make or obtain a list of everyone in the population of interest. Then they draw names or phone numbers, for instance, by chance until the desired sample size is reached (today, most such work is done by computers). Large-scale random sample surveys permit the researcher to draw conclusions about large numbers of people on the basis of a relatively small number of respondents. This is an advantage in terms of time and money. In constructing surveys, sociologists must take care to ensure that the questions and their possible responses will capture the respondents’ points of view. The wording of questions is an important factor; poor wording can produce misleading results, as the following example illustrates. In 1993, an American Jewish Committee/Roper poll was taken to examine public attitudes and beliefs about the Holocaust. To the astonishment of many, results indicated that fully 22% of survey respondents expressed a belief the Holocaust had never happened. Not immediately noticed was the fact that the survey contained some very awkward wording, including the question: “Does it seem possible or does it seem impossible to you that the Nazi extermination of the Jews never happened?” Can you see why such a question might produce a questionable result? The question’s compound structure and double negative wording almost certainly confused many respondents. The American Jewish Committee released a second survey with different wording: “Does it seem possible to you that the Nazi extermination of the Jews never happened, or do you feel certain that it happened?” The results of the second poll were quite different. Only about 1% of respondents thought it was possible the Holocaust never happened, while 8% were unsure (Kagay, 1994). Despite the follow-up poll that corrected the mistaken perception of the previous poll’s results, the new poll was not as method-logically rigorous as it could have been; a single survey question should ask for only one type of response. The American Jewish Committee’s second survey contained a question that attempted to simultaneously gauge two different responses. A weakness of surveys is that they may reveal what people say rather than what they do. Responses are sometimes self-serving, intended to make the interviewee look good in the eyes of the researcher. As we saw in an earlier example, a respondent may not wish to reveal his or her drinking or drug habits.
Fieldwork:
A research method relying on an in-depth and often extended study of a group or community.
Interview:
A detailed conversation designed to obtain in-depth information about a person and his or her activities.
A well-constructed survey, however, can overcome these problems. Assuring the respondent of anonymity, assigning interviewers with whom respondents feel comfortable, and building in questions that ask for the same information in different ways can reduce self-serving bias in survey research.
FIELDWORK
Fieldwork is a method of research that uses in-depth and often extended study to describe and analyze a group or community. Sometimes called ethnography, it takes the researcher into the “field,” directly observing—and sometimes interacting with—subjects in their social environment. Social scientists, including sociologists and anthropologists, have employed fieldwork to study everything from hoboes and working-class gangs in the 1930s (Anderson, 1940; Whyte, 1943) to prostitution and drug use among inner-city women (Maher, 1997) and Vietnam veterans motorcycling across the country to the Vietnam Veterans Memorial in Washington, DC (Michalowski & Dubisch, 2001). Sudhir Venkatesh’s (2006) work on the underground economy is another example of the use of fieldwork in sociological research. Most fieldwork combines several different methods of gathering information. These include interviews, detached observation, and participant observation.
