Sociological Research: Asking and Answering Questions
Asking and Answering Sociological Questions
Introduction to Sociological Inquiry
Sociology today relies on data and evidence to draw conclusions about human behavior.
It employs a range of different methods to study human behavior.
Sociology is considered a science, akin to chemistry and biology.
Initial Question: When asked "Sociology today…"
A: relies on data and evidence to draw conclusions about human behavior.
B: uses a range of different methods to study human behavior.
C: is a science, just as chemistry and biology are sciences.
The correct answer is D: All of the above.
The Impact of Social Media and Big Data
Many individuals use social networking for various purposes, including connecting with friends/family, sharing life details, and forming new relationships.
Platforms like Facebook create detailed records of social relationships.
Scholars utilize data from sites like Twitter and Instagram to analyze social interactions, friendships, and romantic relationships.
Example Study (Backstrom & Kleinberg, 2014):
Claimed Facebook could predict whether romantic relationships would last.
Dataset: Over million individuals in romantic relationships and their Facebook friends (nearly million Facebook users in total).
Finding: If romantic partners had significantly overlapping friendship groups (i.e., the same group of friends), they had a higher likelihood of staying together.
Big Data Potential and Arrogance:
The Internet generates vast amounts of data, enabling sociologists to process more information than ever before, often using statistical or quantitative methods and computer programs.
Concern: Some researchers feel overconfident in claims made from large datasets, leading to an "arrogance of big data."
This can lead to researchers speaking authoritatively about complex social issues (e.g., inequality, poverty) based solely on statistical correlations, without personal involvement or empathy for the studied communities.
Quantitative vs. Qualitative Methods
Sociology maintains a rich tradition that includes both quantitative and qualitative methods.
Qualitative Methods:
Rely on observations, interviews, and archival data.
Emphasize personal involvement and empathy.
Insights often derive from other forms of sociological thinking.
May use data from interactions, conversations, or observations of social scenes.
Definition: Approaches to sociological research that often rely on personal and/or collective interviews, accounts, or observations of a person or situation.
Example: A qualitative study of romantic relationships might use fewer cases to focus on richer details of how individuals dissolve relationships, such as Diane Vaughan's (1986) Uncoupling, which used interviews with people who broke up.
Quantitative Methods:
Make use of numerical data, such as responses collected in surveys or counts of Instagram followers.
Definition: Approaches to sociological research that draw on objective and statistical data and often focus on documenting trends, comparing subgroups, or exploring correlations.
Science in Social Science: Core Values
The scientific validity of sociological research, regardless of method, is defined by a set of values, not sample size or subject matter. These values apply to any empirical research.
Sociological research striving to be scientific adheres to basic standards (King et al., 1994):
Inference: The goal is to generalize beyond specific observations from a particular setting or group to broader phenomena. Sociologists use limited data to make broader claims about what they cannot directly observe.
Example: Diane Vaughan's Uncoupling moved beyond specific interviews to make general claims about how humans form and break romantic relationships, making her analysis scientific.
Reproducibility and Public Procedures: Researchers must make their methods for collecting and analyzing data public, allowing others to retrace their steps, reproduce results, learn from the methods, and assess limitations. This documentation ensures comparability of findings across different researchers and times.
Uncertainty: All scientific conclusions are uncertain. Researchers must be transparent about all sources of uncertainty in their study. High status is conferred upon those who are honest about the certainty of their conclusions.
Reflexivity (Emphasis in Social Sciences): Social scientists must acknowledge that the investigator is an integral part of the world being studied and cannot be completely detached. This includes being aware of power dynamics, and how personal values or identity influence research questions and data interpretation. Researchers must reflect on how their presence and characteristics affect their conclusions.
Example: A middle-class researcher studying a community in poverty should clarify how their class position influenced the sociological argument or relations with subjects.
These four principles represent the highest ideals of social science. While no single study may be strong in all areas, the goal is to achieve as many as possible.
The Seven Stages of the Research Process
Sociological research is best understood as a process, broken down into seven stages:
Define the Research Problem:
All research starts with a problem, which can be factual ignorance about institutions, processes, or cultures.
Best research begins with puzzles: These arise from gaps in understanding, not just lack of information.
Skilled researchers illuminate why events happen, not just what is happening (e.g., "Why are patterns of religious belief changing?" instead of "What are patterns of religious belief today?").
Puzzles can be discovered by reading other research or observing social trends (e.g., shift in attitudes toward mental illness).
Review the Literature:
After identifying a problem, sociologists consult existing research to see if the puzzle has been studied, how others have tried to solve it, and what aspects remain unanalyzed.
This helps clarify issues and appropriate research methods.
Make the Problem Precise:
Clearly formulate the research problem, leading to the development of hypotheses.
Hypotheses: Ideas or educated guesses about a state of affairs, formulated for empirical testing.
The hypothesis must be framed so that collected factual material can either support or disprove it.
Work Out a Design:
Decide how to collect research materials, choosing methods (e.g., surveys, interviews, observational studies) based on study objectives and the specific behaviors being studied.
Carry Out the Research:
Implement the plan from stage . Researchers must be prepared for unforeseen practical difficulties (e.g., inability to contact subjects, non-cooperation from organizations).
Omitting certain persons or institutions due to difficulties could bias results, creating an inaccurate picture.
Interpret the Results:
Analyze the collected data, track trends, and test hypotheses.
Researchers must interpret results to tell a clear story and directly address the initial research puzzle.
Many investigations may not be fully conclusive.
Report the Findings:
Disseminate findings, usually through a journal article or book, detailing the research question, methods, findings, and their implications for social theory, public policy, or practice.
Individual projects are part of a continuous process; reports often suggest new questions for further research.
"Reality Intrudes!": This sequence is a simplified model. Actual research rarely follows these stages neatly; deviations are often necessary and can even be beneficial.
A Historical Contrast: Park and Ogburn
Early sociology was largely theoretical and speculative, but in the , U.S. sociologists at the University of Chicago pushed for grounding theories in facts and data.
This push was embodied by two prominent figures:
Robert Park (1864-1944):
Background: Student of philosophy and a reporter.
Vision: Theories should directly relate to people's actual lives and be based on careful evidence accumulation.
Approach: Advocated for immersive fieldwork ("get the seat of their pants dirty," "wear out their shoe leather"), meeting people in their neighborhoods to understand what was happening.
Methodology: University of Chicago sociologists, following Park, used the city as a laboratory, engaging in participant observation, interviews, and firsthand observations.
Goals: Produced systematic, well-written research reports aimed at improving social conditions (e.g., studying immigration and urban life).
Legacy: Emphasizes personal, emotional, and scientific involvement to develop social world explanations.
William Ogburn (1886-1959):
Vision: Criticized Park's approach, viewing "shoe leather" methods, non-quantifiable findings, and public policy influence as domains of ethics, religion, journalism, or propaganda, not science.
Approach: Argued sociology needed to become a rigorous science focused solely on discovering new knowledge, not making the world better.
Methodology: Advocated for sociology to resemble natural sciences in presentation and orientation, focusing on anything that could be measured with numbers.
Legacy: High value placed on statistics and scientific methodologies, especially with the rise of big data.
Both Park's and Ogburn's visions coexisted and significantly influenced contemporary sociological research, with both qualitative and quantitative approaches having strong foundations.
Research Methods Today: Advantages and Disadvantages
Modern sociology employs a variety of research methods, developing versatile skills in researchers.
1. Ethnography
Definition: The firsthand study of people using observation, in-depth interviewing, or both. Also called fieldwork.
Participant Observation: Researchers directly participate in the activities they are studying.
Other ethnographers may observe from a distance.
Process: The investigator socializes, works, or lives with members of a group/community, explaining and justifying their presence, and gaining cooperation for worthwhile results.
Hazards: Field-workers often experience loneliness, frustration (e.g., when group members refuse to speak frankly), and sometimes physical danger (e.g., researching gangs).
Observer Role: Traditionally, ethnographers aimed for objective reports. More recently, there's increased awareness and discussion of the researcher's connection to subjects, including how their race, class, gender, or power differences might affect the study.
Advantages:
Generates rich, in-depth information on behavior within groups, organizations, and communities.
Provides insights into how people understand their own behavior.
Helps understand both specific groups and broader social processes.
Limitations:
Applicable only to relatively small groups or communities.
Success heavily depends on the researcher's skill in gaining confidence.
Risk of the researcher identifying too closely with the group, losing objectivity.
Potential for conclusions to be influenced by the researcher's own effect on the situation.
Generalizability issues: Findings from one context may not apply to others, and different researchers might draw varying conclusions from the same group.
Recent Example: Forrest Stuart's Down, Out, and Under Arrest: Policing and Everyday Life in Skid Row (2016):
Method: Five years of fieldwork in Los Angeles's Skid Row, observing homeless individuals and policing.
Findings: Documented the powerful impact of the criminal justice system on the lives of homeless people (e.g., Juliette, stopped over times, arrested over times for sitting on the sidewalk).
Insight into policing: Police officers, though not heartless, felt tension between helping the homeless and carrying out arrests. Stuart described their work as "therapeutic policing," using arrest threats to coax people into welfare systems.
Cost of "Therapeutic Policing": Often led to separation of couples/families and loss of possessions if placed in separate shelters.
Policy Implication: Suggested focusing on ample and secure housing for people in financial precarity, rather than solely on policing or social services.
2. Surveys
Definition: Sociological research method in which structured questionnaires are administered to a selected group of people (the population) to gather answers.
Can be administered in person, over the phone, or via mail or email.
Purpose: Best suited for producing less detailed information that is generalizable to a larger population, contrasting with ethnographies' in-depth studies of small slices of life.
Types of Questions:
Standardized (Fixed-Choice) Questions:
Possible Responses: Fixed range (e.g., Yes/No/Don't Know, Very Likely/Likely/Unlikely/Very Unlikely).
Advantages: Easy to count and compare responses statistically.
Disadvantages: May yield restrictive or misleading information by not allowing for subtleties of opinion or expression.
Open-Ended Questions:
Possible Responses: Respondents express views in their own words, allowing researchers to ask follow-up questions.
Advantages: Provides more detailed and nuanced information.
Disadvantages: Lack of standardization makes statistical comparison difficult.
Questionnaire Design: Questions must be clear, follow a set order, accommodate respondents' characteristics (e.g., avoiding unfamiliar terminology, potentially offensive questions, or insufficient information to answer).
Pilot Study: A trial run with a few people to identify and resolve any difficulties before the main survey.
Sampling:
Sample: A small proportion of a larger population (the focus of social research) chosen to represent the whole.
Representative Sample: A sample statistically typical of the larger population, allowing results to be generalized.
Random Sampling: A method ensuring every member of the population has the same probability of being included (e.g., computer-generated random lists, picking every tenth number). This is crucial for large population-based surveys but not always appropriate for specific qualitative populations.
Advantages:
Questionnaire responses are easily quantified and analyzed.
Large numbers of people can be studied efficiently.
Specialized agencies can collect data, given sufficient funds.
Provides statistical measures, aligning with the scientific method.
Disadvantages:
Findings can appear precise but may be dubious due to the shallow nature of many survey responses.
High non-response rates (especially with cell phone reliance), making the representativeness of the sample questionable.
Respondents may perceive surveys as intrusive and time-consuming.
Little is known about the characteristics or views of non-respondents.
Recent Example: General Social Survey (GSS):
Description: Administered to Americans since , often called the "pulse of America." Since , also administered internationally for comparative data.
Significant Finding: An analysis suggested Americans had fewer confidants and an increasing number couldn't name anyone with whom they discussed "important matters." The proportion reporting no confidants jumped from percent in to percent in . The average number of confidants dropped from in to in (McPherson et al., 2006).
Criticism: Some social scientists (e.g., Barry Wellman, 1994) argue these findings don't necessarily prove isolation or loneliness. "Weak" social ties with acquaintances can be rewarding, and overall social ties might be increasing due to the internet, rather than decreasing.
3. Experiments
Definition: A research method used to analyze variables in a controlled and systematic way, either in an artificial situation constructed by the researcher or in a naturally occurring setting.
Considered the best method for ascertaining causality, or the influence of a particular factor on a study's outcome.
Design: Typically involves randomly assigning people to two groups:
Experimental Group: Receives special attention or treatment based on the researcher's theory.
Control Group: Does not receive this special attention.
Subjects are often kept unaware of their group assignment and the experiment's purpose.
Classic Example: Philip Zimbardo's Stanford Prison Experiment (1971):
Method: Set up a mock jail, randomly assigning student volunteers to roles of "guard" or "prisoner."
Aim: Observe how role-playing affects attitudes and behavior.
Results: Guards quickly adopted authoritarian, hostile behaviors, verbally abusing prisoners. Prisoners exhibited apathy and rebelliousness. The experiment was halted early due to high tension and marked effects.
Conclusion: Behavior in prisons is more influenced by the nature of the prison situation itself than by the individual characteristics of those involved.
Advantages:
Allows researchers to test hypotheses under highly controlled conditions.
Disadvantages:
Artificiality: Laboratory settings often fail to replicate natural situations, limiting the generalizability of results to larger society.
Only small groups can be brought into a laboratory.
Subjects knowing they are being studied may behave unnaturally (Hawthorne effect).
Field Experiments: To mitigate artificiality, sociologists sometimes use field experiments, simulating real-life situations as accurately as possible.
Recent Example: Rivera and Tilcsik (2019) Web-Based Experiment on Professor Evaluations:
Hypothesis: College students' evaluations of professors might be affected by gender, specifically influenced by the rating scale used (10-point vs. 6-point, with 10-point potentially triggering "intellectual brilliance" stereotypes associated with men).
Method: Random sample of college students read the same lecture transcript. Half were randomly assigned an instructor with a woman's name, half with a man's name. They were then randomly assigned to evaluate the instructor on either a 6- or 10-point scale.
Results: As hypothesized, male instructors were rated significantly higher than female instructors only when the 10-point scale was used. There were no significant gender differences with the 6-point scale.
Implications: 10-point rating scales in real-world evaluations may yield biased assessments against women.
Statistical Terms
Sociological research often uses statistical techniques. Key concepts include measures of central tendency and correlation coefficients.
Measures of Central Tendency (Calculating Averages):
Example Dataset (Wealth of individuals):
Mean: The average, calculated by adding all values and dividing by the number of cases.
Example:
Advantage: Uses the whole range of data.
Disadvantage: Can be misleading if a few cases are vastly different (e.g., the value skews the mean significantly).
Mode: The figure that occurs most frequently in a dataset.
Example: (appears times).
Disadvantage: Does not account for the overall distribution of data; may not be representative of the whole.
Median: The middle value in a numerically ordered set of figures.
Example: For figures, it's the th figure, which is .
For an even number of figures, it's the mean of the two middle cases.
Disadvantage: Provides no indication of the actual range of the data.
Researchers often use more than one measure of central tendency or calculate the standard deviation (a way of calculating the degree of dispersal, or range, of a set of figures, e.g., from to in the example).
Correlation Coefficients:
Measures the degree to which two (or more) variables are consistently related.
Perfect Positive Correlation: Expressed as (variables correlate completely).
No Relation: Expressed as (no consistent connection).
Perfect Negative Correlation: Expressed as (variables are in a completely inverse relation).
Perfect correlations are rare in social sciences. Coefficients of or more (positive or negative) usually indicate a strong connection (e.g., social class background and voting behavior).
4. Comparative Historical Research
Definition: Research that compares findings from one society with the same type of findings from other societies.
Purpose: Documents how social behavior varies across time, place, and social group membership.
Often quantitative, requiring consistent metrics for comparison of behaviors and attitudes over time and place.
Example: Comparing U.S. divorce rates ( marriages per in ) with those of other countries to assess unique American societal features vs. universal trends.
Historical Research: The most influential type of comparative research.
Classic Study: Theda Skocpol's States and Social Revolutions (1979):
Aim: To develop a theory of revolution based on detailed empirical and comparative historical study.
Method: Analyzed the French, Russian, and Chinese revolutions using various documentary sources.
Findings: Emphasized underlying social structural conditions. Social revolutions are often unintended consequences (e.g., the Bolsheviks did not anticipate the radical transformation of the Russian Revolution).
Contribution: Challenged existing theories that focused mainly on social movements and class relations. Skocpol showed:
State structures are as important as class relations, and more important than the strength of revolutionary movements.
State structures are heavily influenced by international events (e.g., revolutions often follow breakdowns in state authority due to lost international wars).
Contemporary Example: Andreas Wimmer's Waves of War (2012):
Approach: Addresses Eurocentrism in historical sociology by using formal modeling and statistical techniques to analyze hundreds of cases simultaneously (rather than a few famous ones), creating large original datasets.
Thesis: What happened in Haiti or Latin America is as important as events in Europe.
Findings: Identified a major shift in the drivers of war throughout history.
Pre-19th Century: Most wars driven by conquest or balancing power.
More Recently: Wars are increasingly driven by ethnic and nationalist concerns.
Conclusion: Challenged Karl Marx's prediction of 20th-century class struggle, arguing it became the age of ethno-nationalist conflict.
Key Statistical Finding: The presence of nationalist organizations in a territory more than doubles the probability of war at any given time. This finding would be difficult to obtain with small-case qualitative approaches.
Unanswered Questions and Ethical Considerations
Can Sociology Identify Causes and Effects?
A primary challenge in research methodology is distinguishing cause and effect.
Difficulty: It is hard to determine if one social context produces an effect, or if certain types of people (who might experience that effect) tend to be drawn to or can only afford to live in that context.
Example: Is living in a low-income neighborhood the cause of unemployment or health problems like diabetes, or do unemployed or unhealthy people tend to reside in such neighborhoods due to affordability?
There is a growing "crisis of confidence" among scholars regarding the ease of achieving clear cause-and-effect identification.
How Can Social Research Avoid Exploitation?
All research involving human beings presents ethical dilemmas, especially concerning risks to subjects.
Key Question: Does the research pose risks to subjects that are greater than the risks they face in their everyday lives?
Example: Ethnographers in high-crime areas could risk their subjects' arrest through their writings or risk their own arrest simply by observing.
Exploitation Concern: Are social scientists benefiting at their subjects' expense? (e.g., should profits from a book based on research subjects' cooperation be shared with them?)
This question of exploitation is more pronounced in qualitative field studies.
Community-Based Participatory Research (CBPR):
Definition: A collaborative approach to research that equitably involves community members, researchers, and others in all aspects of the research process.
Principle: Adheres to "nothing about us, without us," recognizing that the voices and distinctive strengths of those being studied must be respected and influence the research question, design, and implementation.
Implication: Communities affected by research have a right to influence what is researched and how it is done (Wilson et al., 2018).
Can We Really Study Human Social Life in a Scientific Way?
Definition of Science: The use of systematic methods of empirical investigation, the analysis of data, theoretical thinking, and the logical assessment of arguments to develop a body of knowledge about specific subject matter.
Conclusion: Based on this definition, sociology is indeed a scientific endeavor.
Differences from Natural Sciences:
Humans are self-aware beings who assign meaning and purpose to their actions, unlike natural phenomena or animals.
To accurately describe social life, sociologists must grasp the concepts people apply to their own behavior (e.g., understanding a death as a suicide requires knowing the individual's intent).
Advantages for Sociologists: Researchers can directly pose questions to their human subjects, which is not possible in natural sciences.
Unique Difficulties for Sociologists:
People aware of being scrutinized may behave unnaturally.
Subjects may consciously or unconsciously portray themselves differently from their usual attitudes.
They may try to "assist" the researcher by giving responses they believe are desired.
Reading a Table (Practical Guide)
Tables condense information and are crucial for understanding and verifying sociological conclusions.
Read the title in full: Understand the subject, comparison, and limitations of the data (e.g., "Opinion of the United States: Comparison of Selected Nations" indicates specific data across countries over time).
Look for explanatory comments or notes: Check source notes (e.g., "Pew Research Center, 2022c") and footnotes (e.g., "Data not available for all nations for all years") for details on data collection, reliability, and original sources.
Read the headings: Identify what information is contained in each row (left-hand side, e.g., Countries) and column (top, e.g., Years and Proportion of Population with "Favorable" Opinion).
Identify the units used: Determine if figures represent cases, percentages, averages, etc. Convert figures if needed for better understanding (e.g., converting actual numbers to percentages).
Consider the conclusions: Analyze what the data suggests, both in line with the author's discussion and by formulating further questions or issues that the data might raise (e.g., explaining sudden drops in favorable opinion toward the U.S. in certain countries).
```