Research Process in Sociology

Scientific Method in Sociology

Sociologists use theory as a framework and follow the scientific method to research particular questions about society and social behavior.

Sociologists use several different research methods to study society and social behavior. Their research questions are guided or influenced by theory. A theory is a set of arguments that seeks to explain or predict a particular aspect of social life. Major theories used in sociology include conflict theory, functionalism, symbolic interactionism, feminist theory, and intersectionality. Conflict theory is based on the idea that society is characterized by conflict and competition between different social groups. Functionalism seeks to explain how different parts of society work together to create order. Symbolic interactionism emphasizes the interactions between individuals and society. Feminist theory analyzes how gender norms and gender inequality impact society and social behavior. Intersectionality considers how gender, race, class, age, ability, and other factors of identity combine to create different experiences of social stratification. Sociologists consider the arguments and concepts of one or more theories to develop research questions and research proposals.

Guided by a theoretical stance, researchers design studies to collect data about society and social behavior. Sociology is a science and follows the scientific method to conduct research. In using the scientific method, sociologists aim to gather accurate, reliable data and to minimize the potential for bias or subjectivity in their research. Sociologists use various research methods to collect and analyze data, such as surveys, interviews, and statistical analysis. Each research method is grounded in the scientific method.

Prior to considering research methods, it is important that researchers develop a strong research question. Sociologists then select a research method according to the question they are trying to answer. This is an important step because certain methods work best for certain types of questions. For example, researchers studying representations of gender in the media might take different approaches depending on the particular question they want to investigate. To analyze changing trends in the representation of gender on television, a researcher might analyze data about numbers of characters who conform to particular gender norms. To analyze the impact of a particular representation of gender, a researcher might use a survey or interview to collect data about viewer reactions. However, all studies begin with a research question, informed by observations. Researchers follow the scientific method to investigate the question and then draw conclusions based on data.

Research Steps

Sociologists follow the scientific method to conduct research and report results. Some studies use qualitative data, while others use quantitative data.

Research Steps in Sociology

Research steps include identifying a topic or question, reviewing existing research, forming a hypothesis or argument, selecting a research method, collecting and analyzing data, drawing conclusions, and reporting results.

Identify a Topic

The research process begins when sociologists identify a question or issue that warrants investigation. Researchers first make observations about a particular issue, based on their academic interests and area of expertise. These observations, along with a particular theoretical framework, guide researchers to formulate a particular question to explore. For example, many sociologists are interested in gender. But there are many questions to research about gender: how is gender embodied; how do gender differences play out in an educational institution, the economy, religion, or family; how do media reflect ideas about gender? Sociologists make observations about the issues of interest, informed by a theoretical approach, and develop a specific question. For instance, a researcher interested in the role of gender in education might want to investigate whether the experiences of male and female graduate students in the sciences are similar or different.

Review Existing Research

The next step in the research process is to conduct a literature review. Sociologists review available research related to the topic under consideration to find out what other researchers have reported. They use peer-reviewed research that has been published in scholarly journals. Peer-reviewed studies are published after other sociologists who were not involved with a study carefully review the methods, data, and conclusions of a study. Scholarly journals are reputable academic journals that publish articles that have passed the peer-review process. Reviewing the research and articles that other sociologists have already published provides a base of information that researchers can use and build upon. Previous studies often highlight areas and questions that need further exploration. For example, a researcher who wants to study whether race impacts attitudes about vaccination will first search for peer-reviewed studies related to this topic. The researcher might find that previous studies do not sufficiently address a particular issue related to race and vaccination. This can help the researcher frame a specific question to address in a new study.

Form a Hypothesis or Argument

After reviewing current research, sociologists form a hypothesis, a predictive statement about the relationship between an independent variable and a dependent variable. A variable is a concept or feature with multiple values (such as gender, race, social class, age, etc.). An independent variable is a concept or feature that causes change in another variable. The dependent variable is affected or changed by the independent variable. This variable is the outcome the researcher is trying to explain. In sociology, researchers must define or conceptualize the particular variables under consideration. For example, a researcher who studies gender as a variable must first conceptualize gender, defining what gender means or entails.

In social science research, it is not always necessary to form a hypothesis prior to conducting research. Some social science research takes an inductive approach. This entails doing research on an issue or question and then interpreting the research findings in order to form a hypothesis or argument. One example of this type of research is grounded theory, an approach to research that begins with the collection of qualitative data (nonnumerical data). Researchers then look for patterns in the data and form an argument based on these patterns. For instance, a researcher interested in how people engage with a health care system might conduct in-depth interviews with members of the population being studied. The researchers would then look for patterns in the responses to form an argument about how that group uses health care or feels about the health care system.

Select a Research Method and Collect Data

Sociologists use a number of different methods to collect data. Researchers consider what method will best answer the research question under consideration and allow them to accept or reject the hypothesis or argument they have formulated. There are two major types of sociological research: quantitative research, which is number-based research, and qualitative research, which is an approach that uses nonnumerical data. Both types of research involve careful observation and interpretation. Quantitative data is sometimes considered more objective than qualitative data. An example of quantitative research is Seth Stephens-Davidowitz's use of data collected from websites to determine trends in various types of social behavior. By counting things like the number of times people searched certain keywords or clicked on certain links, Stephens-Davidowitz gathered information about people’s habits, interests, and beliefs. He then drew conclusions, based on patterns in the data. Qualitative research takes a different approach. This type of research focuses on rich, deep meaning and seeks to go beneath the surface of data. In this way, it can tell more of the social story in question than quantitative data. A well-known example of qualitative research is Arlie Hochschild's study of gender roles based on interviews with married couples conducted over a span of eight years. For both qualitative and quantitative research, there are several methods that sociologists use, including surveys, interviews, and statistical analysis. Researchers choose an approach and then design a study in order to collect data.

Analyze Data

After data have been collected, the next step is to analyze the data. Quantitative data are analyzed through sophisticated computer programs that allow for multiple variables to be examined simultaneously. Qualitative data can also be analyzed through computer software that codes the data by evaluating words, phrases, and themes collected through research. There are numerous statistical techniques that can be used to analyze both quantitative and qualitative data. Researchers look for correlations between the data (relationships between two or more variables). The strength of the relationship reveals the degree to which one variable causes or is linked to change in another variable. For example, a researcher might collect data about the amount of time people spend on housework. To analyze these data, the researcher might look at the strength of the relationship between gender and the amount of time people spend doing housework.

Draw Conclusions and Report Results

Finally, in the last stage of the scientific method, the researcher draws conclusions from the study. For example, in studying the relationship between gender and housework, a researcher might draw conclusions about what social forces lead to women doing more housework than men. Often, researchers refine their original hypothesis or argument, based on the conclusions they draw from their research. If a quantitative study has been conducted with a large enough sample size, results may be generalized to a larger population. Qualitative studies are more likely to give rich, detailed information about the specific group studied and cannot be generalized. In reporting results, researchers consider the validity and reliability of their study. Validity is the extent to which a study measures what it is intended to measure. Reliability is the likelihood that if a study is repeated the researcher will obtain consistent results. Studies that are valid and reliable are accepted by other researchers. Researchers usually publish their results as articles in peer-reviewed journals. These are academic journals that publish papers that have been reviewed and accepted by other researchers with expertise in the same subject under consideration in an article. In the case of sociology, other sociologists who were not involved with the study review the article. If the reviewers agree that the study is valid and reliable and that it adds important information to the field of sociology, they recommend that the article be published.

Qualitative Research in Sociology

Qualitative research methods include survey research, interviews, participant observation, content analysis, and comparative-historical analysis.

Qualitative research is an approach that uses nonnumerical data, such as analysis of interview responses or observed behavior, to study the social world to study the social world. Examples of this type of research include the analysis of interview responses or of observed behavior. Qualitative research is often characterized by questions that seek to describe a process or understand the experience of a person or small group of people. Qualitative methods can use inductive reasoning, the process of reaching a conclusion by starting with specific observations and drawing general conclusions.

Qualitative data provides an analysis that goes beneath the surface of data. It generates rich, deep meaning and usually tells the researcher more about the individual, group, or process in question. Written responses or verbal interviews are common examples of qualitative data. A key part of qualitative analysis is identifying the crucial aspects of a study in order to apply them to a theory or pattern. To do so, researchers use a system to code the information in the data they collect. This allows the researchers to look for patterns and themes and then create categories that can be used in quantitative ways. For example, a researcher might want to study students' commuting experience. To collect qualitative data, the researcher can ask students to write a few sentences explaining how they got to school, thus describing the overall experience in detail. The researcher then codes the responses, looking for patterns. The researcher may look for certain types of language that describe commuting as burdensome or easy, safe or unsafe, affordable or unaffordable, etc. The researcher flags (codes) answers that fall into the categories being studied. This coding can be done by hand, for example, by highlighting each type of response in a different color. Then the number of answers in each category can be counted. Many researchers use coding software to speed up the process. The coding gives the researcher information about the prevalence of certain attitudes, obstacles, or issues related to the subject of student commutes.

One issue with qualitative data is the difficulty of generalizing findings to aggregate groups (such as a whole society). Qualitative research is more subjective as it relies on the lived experiences of a small group of people. The planning done for qualitative research is typically less intensive in the beginning than for quantitative research. However, the data analysis phase is much more time-consuming. Because qualitative research seeks to understand a small number of people in great depth, the results may not apply to as many individuals or groups as aggregate, quantitative data. This does not mean that qualitative methods are not valid or reliable. A study is valid if it measures what it is intended to measure. Reliability means that if the study is replicated (done again), it shows the same results. Qualitative research can be valid and reliable, but researcher bias is more common in qualitative than quantitative research. All researchers must be careful to consider their own biases––their perspectives, beliefs, values, and expectations. These biases can shape what types of questions are studied, as well as how answers and other data are interpreted.

Quantitative Research in Sociology

Quantitative research emphasizes objective measurements and mathematical, or numerical, analysis of data.

Quantitative research is an approach that uses numerical data, such as percentages and rates, to study the social world. Quantitative research is often characterized by questions about what factor might affect another factor. For example, it might be used to try to determine if gender affects salaries. Quantitative research aims to determine the relationship or correlation between two variables. This type of research uses deductive reasoning, the process of reaching a conclusion by starting with a general hypothesis and generating a specific idea or argument. Deductive reasoning is used when the researcher already has an idea of what data might show or whether a particular theory is appropriate or applicable. Researchers gather data and analyze them to determine whether the data support or refute their original hypothesis. Quantitative methods are more objective than qualitative methods. However, they do not always provide all necessary or important information. For instance, to study students' commuting experience, a researcher could provide a form with questions and check boxes with many possible answers. One question could ask what mode of transportation is used, another could ask how much money students spend on commuting per month, and so on. This method of data collection provides specific information. The information is easily quantifiable, or countable. The researcher can determine what percentage of respondents use cars, ride bicycles, take buses, and walk, as well as how much the average student spends. A great deal of information can be collected this way. However, the researcher must still interpret the data in order to draw conclusions.

Qualitative Research Examples	Quantitative Research Examples
The Second Shift by Arlie Hochschild This book is based on intensive interviews with 50 couples over eight years, to learn about roles and routines. It was published in 1989, with an update in 2012. Off the Books by Sudhir Venkatesh This book uses participant observation of drug dealers, prostitutes, and their surrounding neighborhood on the south side of Chicago during the late 1990s and early 2000s. In participant observation, researchers become part of the group they are studying. They join a community in order to understand it from the inside. This book was published in 2006. Unequal Childhoods by Annette Lareau This book is based on participant observation of 12 African American and white families. Lareau followed up on the study 10 years later, returning to conduct interviews with the families. It was first published in 2003; the revised book was published in 2011.	Everybody Lies by Seth Stephens-Davidowitz This book is based on analysis of a huge amount of data collected from Google Trends, Google AdWords, Wikipedia, Facebook, and various other sites. The author uses the data collected to examine various types of social behavior, including how parents treat boys and girls, expressions of racism, and reading habits. It was published in 2017. "Is There a Male Marital Wage Premium? New Evidence from the United States" by Volker Ludwig and Josef Brüderl This article uses an analysis of data from the National Longitudinal Survey of Youth (1979–2012) to consider findings from other studies about why married men earn more than unmarried men. This study replicates findings from other studies but argues that there is no causal connection between marriage and earnings. It was published in 2018.

Qualitative Research
Examples

Quantitative Research
Examples

The Second Shift by Arlie Hochschild
This book is based on intensive interviews with 50 couples over eight years, to learn about roles and routines. It was published in 1989, with an update in 2012.

Off the Books by Sudhir Venkatesh
This book uses participant observation of drug dealers, prostitutes, and their surrounding neighborhood on the south side of Chicago during the late 1990s and early 2000s. In participant observation, researchers become part of the group they are studying. They join a community in order to understand it from the inside. This book was published in 2006.

Unequal Childhoods by Annette Lareau
This book is based on participant observation of 12 African American and white families. Lareau followed up on the study 10 years later, returning to conduct interviews with the families. It was first published in 2003; the revised book was published in 2011.

Everybody Lies by Seth Stephens-Davidowitz
This book is based on analysis of a huge amount of data collected from Google Trends, Google AdWords, Wikipedia, Facebook, and various other sites. The author uses the data collected to examine various types of social behavior, including how parents treat boys and girls, expressions of racism, and reading habits. It was published in 2017.

"Is There a Male Marital Wage Premium? New Evidence from the United States" by Volker Ludwig and Josef Brüderl
This article uses an analysis of data from the National Longitudinal Survey of Youth (1979–2012) to consider findings from other studies about why married men earn more than unmarried men. This study replicates findings from other studies but argues that there is no causal connection between marriage and earnings. It was published in 2018.

<Vocabulary >Researcher Bias and Rapport with Study Subjects

Research Methods

Contents