Funding decisions will be communicated to applicants by the end of August The successful grant must start by 15 November Toggle navigation. Home Funding Funding opportunities Guidance for applicants Guidance for grant holders Guidance for large investments Guidance for peer reviewers Research Our research Research topics Future of social science Impact toolkit International research Celebrating Impact Prize Research and impact evaluation Collaboration Collaboration opportunities Opportunities for business Working with policymakers Opportunities for civil society Guidance for collaboration Postgraduate collaboration Skills and careers Doctoral training Postgraduate careers Writing competition Media training Public engagement Festival of Social Science Social Science for Schools Photographic competition Public engagement guidance Public dialogues News, events and publications News Events Publications Impact case studies Evidence briefings About us What we do Governance and structure Strategy and priorities Policies and standards Performance information Visual identity and logos What is social science?
What might the boss expect you to do if you do not get a raise? Can the boss replace you with someone as good for the same or less money? What kind of bargaining power do you and other employees have? There are many rival explanations for why you did not get a raise, only one of which consists of the boss's appraisal of your work.
Thinking like a researcher can help you assess the evidence for inferring that one factor caused another. The strongest way to rule out all rival explanations is to conduct a tightly controlled experiment where subjects are randomly assigned to two groups, only one of which is exposed to the independent variable while the other is not.
Approaches to Understanding and Measuring Wellbeing
Happily, no experimenter has the power to assign us randomly to groups and then tell us how we must or must not find a job so that the effect on our income can be studied. Many areas of sociological research share these ethical and practical constraints. In such situations we can only try to approximate the logic of experimental designs by controlling for as many rival explanations as possible.
Steps in the Research Process. Although not all research studies follow the same pattern, it is possible to spell out the steps that occur frequently in the research process.
Download Product Flyer
Defining the problem involves selecting a general topic for research, identifying a research question to be answered, and defining the concepts of interest. Individuals have personal research questions, and social researchers have more general ones. You may wonder, for example, how you will get your first job. On a larger scale, sociologists might ask how people in general find jobs, as Granovetter did The next step is to review the existing literature to determine what is already known about the problem.
Prior work may offer general descriptions, raise some key questions, discuss the strengths and limitations of measures that have already been tried, and suggest profitable lines of further research. More and more libraries offer computerized literature searches that speed up the review process. Ideally, in their effort to build knowledge, researchers develop several competing hypotheses. Durkheim did this in his classic study entitled Suicide.
He considered the possibility that suicide rates varied as a result of heredity, climate, or social factors. He found social factors, such as the presence or absence of cohesion within a social group, to be the most important determinant of suicide. Researchers then decide on a design for the study that will allow them to eliminate one or more of the hypotheses. Research design is the specific plan for selecting the unit of analysis; determining how the key variables will be measured; selecting a sample of cases; assessing sources of information; and obtaining data to test correlation, establish time order, and rule out rival hypotheses.
Sociologists gather information in a variety of ways, depending on what they want to investigate and what is available. They may use field observations, interviews, written questionnaires, existing statistics, historical documents, content analysis, or artifactual data. Each of these methods will be discussed briefly in the next section. Once the data are collected, they must be classified and the proposed relationships analyzed. Is a change in the independent variable indeed related to a change in the dependent variable? Can time order be established?
Are alternative explanations ruled out? Drawing conclusions involves trying to answer such questions as these: Which of the competing hypotheses are best supported by the evidence? Which are not? What limitations in the study should be considered in evaluating the results? What lines of further research does the study suggest? Conclusions rest heavily on the way research is designed and data are gathered. Social researchers study and try to understand the social world.
Either they seek to describe some feature of social life or they try to analyze and explain interrelationships among social factors. Various types of data are available for both goals, and those data may be collected in different ways. Does early childhood education for children living in poverty help them to succeed in school and beyond? In a social experiment, the researcher tries to see whether a change in the independent variable in this case, exposure to a preschool program is related to a change in the dependent variable school success or failure, criminal arrests, teen pregnancies, unemployment, and the need for welfare , while other conditions are held constant family and neighborhood.
In an experimental design, the effect of the independent variable is assessed by comparing two groups of people. One group, the experimental group, is exposed to the hypothesized independent variable the preschool program , while another group, the control group, is not. To rule out other explanations, the experimental and control groups must be identical in every respect except their exposure to the treatment. In the Perry preschool study there was an experimental group of 58 and a control group of They were selected for the study at age 3 or 4 on the basis of parents' low educational and occupational status, family size, and children's low IQ intelligence test scores.
Pairs of children matched on IQ, family socioeconomic status, and gender were split between the two groups. The experimental group attended a preschool program for two years. Studies of the Perry Preschool Program in Michigan and five other preschool programs show significant differences between children in the experimental and control groups in terms of their higher intellectual performance as they began elementary school, their lesser need to repeat a grade or to receive special education, and their lower rates of dropping out of high school.
In the Perry preschool study, the two groups were also compared in their early adult lives.
Nineteen-year olds who had attended the program were better off in a variety of ways than the control group. The program seems to have increased the percentage of participants who were literate from 38 to 61 percent , enrolled in postsecondary education from 21 to 38 percent , and employed from 32 to 50 percent.
Moreover, the program seems to have reduced the percentage of participants who were classified as mentally retarded during their school years from 35 to 15 percent , school dropouts from 51 to 33 percent , pregnant as teenagers from 67 to 48 percent , on welfare from 32 to 18 percent , or arrested from 51 to 31 percent Schweinhart and Weikart, , pp.
A social science degree can provide some answers – or at least some perspectives.
Experiments are strong methods for meeting the three criteria of time order, correlation, and the elimination of rival hypotheses needed for inferring causality. They are limited by the practical and ethical restraints that exclude the study of private or dangerous behavior. Another method--interviews--can help researchers to obtain information about private, personal, or taboo attitudes and behaviors.
What kind of gender-role behavior occurs between long-term partners in a relationship?
Are there differences in the gender roles people assume when couples are straight heterosexual and gay composed of two homosexual men or lesbian women? These are some of the research questions posed by Philip Blumstein and Pepper Schwartz , two sociologists at the University of Washington. To investigate these and related issues, they conducted interviews with more than six hundred people living in long-term relationships, and they mailed a written questionnaire to more than ten thousand people who agreed to participate in the study.
See Table 2. Although carefully protecting the identities of the individuals involved, the researchers collected background information on the respondents' educations, ccupations, incomes, and ethnicity, as well as considerable information about their relationships with their partners. The use of interviews and questionnaires enabled them to ask everyone the same questions, so that comparisons could be made between long-term and short-term couples; between gay, lesbian, and straight couples; between couples with children and those without; and so forth.
Practical and ethical considerations would have made it impossible to gather such data by observation, and other methods of data collection would have been equally inappropriate. Surveys are useful for describing the characteristics of large numbers of people in an efficient way. In this case, if only a few individuals had been studied, we might think that the results were unique to them and did not occur in the larger population.
Surveys of carefully selected samples permit the accurate determination of rates of behavior or the frequency with which certain attitudes are held. The special sampling procedures researchers have developed are among the most powerful tools in their kit.
- Social Science: Helping Us Understand Ourselves and the World!
- Knowledge, Truth, and Social Reality: An Introductory Note on Qualitative Research!
- Games that boost performance : [30 ready-to-use group activities]!
- THE UK HOUSEHOLD LONGITUDINAL STUDY.
- Social Sciences.
- Describing How.
Properly done, sampling permits conclusions about entire populations of individuals, groups, organizations, or other aggregates by studying only a few of them. The key lies in how those few are selected.
Understanding Social Research | SAGE Publications Ltd
A population is the total number of cases with a particular characteristic. Suppose you were interested in the sexual attitudes of American college students. Do you think you could walk out the door wherever you are and select the first ten warm bodies you encountered, interview them, and draw accurate conclusions about the attitudes of all college students?
Such a technique is likely to be very unrepresentative. To overcome this problem, researchers use random sampling. There are many types of scientific samples. In a random sample, every element person, group, organization, or whatever of the population must have an equal and known chance of being selected for inclusion in the sample. There is solid technical knowledge available about sampling, but we cannot cover it all in an introductory sociology text.