1. Chapter 2: Process and Problems of Social Research
Question?
Strategy?
Theory?
Design?
Ethical?
The text presents this information in a slightly different order.
The type of research question will determine the research strategy, deductive, inductive or descriptive, whether or not you will be testing a theory, and if your study is a cross-sectional or longitudinal design.

2. The Question
Do mandatory arrest policies decrease domestic violence recidivism?
Does how much a person loves his or her dating partner influence the frequency of lying?
What factors influence men’s and women’s level of retirement income?
From Text:
Beginning in 1981, The Police Foundation and the Minneapolis Police Department began an experiment to find the answer. The Minneapolis experiment was first and foremost scientifically relevant: it built on a substantial body of contradictory theorizing about the impact of punishment on criminality (Sherman & Berk, 1984).

3. What Makes a Question “Good”?
1. Feasibility
Resources
Time
From Text:
Many social scientists evaluate their research questions in terms of three criteria: feasibility given the time and resources available, social importance, and scientific relevance (King, Keohane, & Verba, 1994):

4. 2. Social Importance
Make a difference
Help understanding

5. 3. Scientific Relevance Resolve contradictory research findings
Modify social theory? *

6. Deductive, Inductive, or Descriptive?
The Strategy
The Circle of Science includes the following components:
Theory A logically interrelated set of propositions about empirical reality.
Hypothesis A tentative statement about empirical reality, involving a relationship between two or more variables.
Variable A characteristic or property that can vary (take on different values or attributes).
Independent variable A variable that is hypothesized to cause, or lead to, variation in another variable.
Dependent variable A variable that is hypothesized to vary depending on or under the influence of another variable.
Data include the systematically gathered evidence. Science requires evidence to be obtained empirically—through the senses. In the social sciences this usually means through “hearing” respondents’ answers to survey or interview questions or through “observing” behavior or artifacts.
Empirical generalizations are the organized patterns we observe in the data. We formulate these generalizations through analyses, qualitative or quantitative.
Deductive, Inductive, or Descriptive?

7. Strategy #1: Deductive Research
Deductive research starts with a theory and deduces a specific expectation.
From text:
deductive research starts with a theoretical premise and deduces a specific expectation.

8. Strategy #2: Inductive Research Inductive Research starts with data.
From Text:
In contrast to deductive research, inductive research begins with specific data, which are then used to develop (“induce”) a theory to account for the data. One way to think of this process is in terms of the research circle: rather than starting at the top of the circle with a theory, the researcher starts at the bottom of the circle with data and then develops the theory.

9. Inductive Research
If police noticed that arresting at least one person from the scene of a domestic violence case resulted in fewer subsequent calls to that same scene,
The police officers might develop a theory that mandatory arrests decrease domestic violence recidivism. *
Using the inductive research strategy to understand the effects of mandatory arrests would require us to start with data already collected on this phenomenon.

10. Strategy #3: Descriptive Research
Descriptive research starts with data and proceeds only to making generalizations, not generating entire theories.
From Text:
Both deductive and inductive research connect theory with data. Descriptive research by itself does not, but is still part of the research circle. As this graphic indicates, descriptive research starts with data and proceeds only to the stage of making empirical generalizations, not generating entire theories.
Valid description is actually critical in all research. Much important research for the government and private organizations is primarily descriptive: How many poor people live in this community? Is the health of the elderly improving? How frequently do convicted criminals return to crime? Description of social phenomena can stimulate more ambitious deductive and inductive research. Simply put, good description of data is the cornerstone for the scientific research process and an essential component for understanding the social world.

11. Descriptive Research
“How frequently do those arrested for domestic violence return to violence?”
Description of social phenomena can stimulate more deductive and inductive research
Good descriptions of data are the basis for scientific research process and
An essential component for understanding the social world. *
Here’s how a descriptive strategy might be applied with our example of mandatory arrests in domestic violence cases.

12. The Theory
Social theory: A logically interrelated set of propositions about empirical reality
(i.e., the social world as it exists)
Theories and Research
Make connections to:
Social processes (reality)
Large bodies of research
Creating and evaluating theory
Ask student to explain what a theory is. Then present this information.

13. Example of Theory

14. What Is the Design? Cross-sectional Longitudinal Once you’ve chosen a:
Research question,
Strategy, and
Theory
Decide on a research design
Cross-sectional
Longitudinal
In addition to the question strategy and theory, a researcher has to decide on the design. Will you study a cross-section, a snapshot of the phenomenon of interest? Or will you observe changes over time?

15. Cross-Sectional Design
Data collected at one point in time.

16. Cross-Sectional Design

17. Cross-Sectional Design Application
To study the effects of mandatory arrests
Take “cross-section” of all domestic violence cases and
Examine whether arrests affected recidivism*

18. Strengths of Cross-Sectional Design
Relatively quick and easy to conduct
Data collected on different kinds of people in relatively short time.
Good for descriptive analyses
Good for generating hypotheses.
Large samples enable use of inferential statistics (e.g. to compare subgroups within the sample).

19. Weaknesses of Cross-Sectional Design
Difficult to establish time order
Unable to chart individual variations in development or changes
Unable to chart changing social processes over time.

20. Longitudinal Designs

21. Longitudinal Designs: The Panel Design
A sample (called a panel) is drawn from a population at time 1, and data are collected from the sample (for instance, 100 arrestees from domestic violence cases are selected and interviewed).
2. Time passes, some panel members become unavailable for follow-up (move or refuse to continue participating).
3. At time 2, data are collected from the same people (the panel) as at time 1—except for people who cannot be located (the remaining arrestees are re-interviewed).

22. Weaknesses of Panel Designs
Expense and Attrition
Expensive to keep track of individuals for long periods of time
Panel members who can be located declines over time.
Subject fatigue
Weary of repeated interviews and drop out
Used to answering questions--Start giving stock answers
Expensive and attrition - Miller 1991: 170
Subject fatigue – Campbell 1992

23. Longitudinal Designs: Repeated Cross-Sectional Design (Trend study)
1. A sample of domestic violence cases is drawn from a population of cases at time 1, and data are collected from the sample.
2. As time passes, some people leave the population and others enter it.
3. At time 2 a different sample of cases is drawn from this population.

24. Longitudinal Designs: A Cohort Study
Cases in Cases in
In a cohort study, the follow-up samples (at one or more times) are selected from the same cohort—
Cohort: People who have experienced a similar event or a common starting point
(Example: Baby Boomers, Gen-Xers)
Finally, another design, the cohort study. A cohort
From Text:
Or the population may be defined narrowly, as when members of a specific age group are sampled at multiple points in time (a cohort study). The frequency of follow-up measurement can vary, ranging from a before-and-after design with just one ­follow-up to studies in which various indicators are measured every month for many years.
Cohort:: Individuals or groups with a common starting point. Examples include college class of 1997, people who graduated from high school in the 1980s, General Motors employees who started work between 1990 and the year 2000, and people who were born in the late 1940s or the 1950s (the “baby boom generation”).

25. Example of a Cohort

26. Real-life Example
Data collected from high school students from Baltimore, Maryland
Studies the difference in initiation of tobacco use between a cohort of adolescents that started working for pay and a cohort of adolescents that did not work.
The results suggest that adolescents who work for pay have a higher risk of initiating tobacco use.
Ramchand, R., Ialongo, N. S., & Chilcoat, H. D. (2007). The effect of working for pay on adolescent tobacco use. American Journal of Public Health, 97(11),

27. Unit of Analysis
Major entity you are analyzing in study. Any of the following could be a unit of analysis:
Individuals
Groups
Artifacts (books, photos, newspapers)
Geographical units (town, census tract, state)
Social interactions (dyadic relations, divorces, arrests)

28. Units of Analysis and the Domestic Violence Example
Cases = individual arrests
Cases =precincts
Precinct #1
Precinct #2
Precinct #3
From Text:
There is one final concept you need to understand in planning the overall design of your project: “the unit of analysis.” Most sociological and psychological studies gather data from, and formulate conclusions about, individuals—so individuals are called the units of analysis. The researcher may collect survey data from individuals, analyze the data, and then report on, say, how many individuals felt socially isolated and whether drug abuse by individuals was related to their feelings of social isolation. Both data and conclusions would be about individuals, not groups. One could also then generalize about individuals; for example, perhaps you find that women are less likely to be isolated than men. The finding talks about many people (women vs. men), but still points to a phenomenon occurring within individuals—that is, that individual women are less likely to be isolated then individual men. Individuals remain the unit of analysis.
But some research gathers data about groups of some sort, such as families, schools, work organizations, towns, states, or countries. In this case, groups are the units of analysis. One may find, for instance, that smaller high schools have lower levels of student violence than large schools. In this case, the finding is not about students, but about certain kinds of schools. Here, the organization, not the individual, is the unit of analysis. It’s important to recognize your units of analysis, in order to avoid logical errors in your analysis. In Chapter 5, we’ll discuss the logical errors that can result when we use aggregate data
Individuals as Unit of Analysis
Groups as Unit of Analysis

29. Two Research Fallacies
A fallacy is an error in reasoning
Ecological fallacy occurs when you make conclusions about individuals based only on analyses of group data.
You measured math scores of a class and found that they had the highest average score in the district.
Later you meet one of the kids from that class and you think "she must be a math genius.”
Fallacy! Just because she comes from the class with the highest average doesn't mean that she does well in math.
She could be the lowest math scorer in a class that otherwise consists of math geniuses!

30. Two Research Fallacies
An Exception fallacy is sort of the reverse of the ecological fallacy.
Occurs when you reach a group conclusion on the basis of exceptional cases.
This fallacious reasoning is at the core of a lot of sexism and racism (stereotypes).
The man who sees a woman make a driving error and concludes that "women are terrible drivers.
Fallacy!

31. Errors in Reasoning

32. Ethical Considerations
Honesty and Openness
Should not conceal motives to subject
Full disclosure of researcher identity
The Uses of Science
How will findings be used?
Who controls final report?
Who controls publicity?
Confidentiality
Informed consent agreement
Identifying information is available only to designated research personnel for specific research needs
Institutional Review Board (IRB)
Federally funded research

33. Conclusion: The Process of Researching the Effects of Mandatory Arrests

34. For Your Own Research… Contemplate: the question, the strategy,
the theory,
the design, and
the ethics of the project.

35. Populations and Samples