9/12/13 – Thursday
How Sociologists Do Research
• The approaches and methods used by social science researchers
• Science is a logical system that bases knowledge on direct systematic observation
What is a Valid Sociological Topic?
• Sociologists conduct research on almost every area of human behavior.
This includes research at both the macro level and micro level
Sociological research includes a specific method of how it is to be done.
Common Sense and the Need for Sociological Research
• Sociological research is needed because common sense is highly limited as a source of
knowledge and its insights are often suspect.
Eight Steps of a Research Model
Step 1: Select a topic
• Asociologist may choose a particular topic for a variety of reasons. For example, he or
she may be interested in the topic, there may be funding available to do research on the
topic, and/or the topic may constitute a pressing social problem that the sociologist wants
to help people better understand and, perhaps, help to solve.
Ways to Select Topics
• Personal Experience
• Curiosity based on media
• State of knowledge in the field
• Solving a problem
Step 2: Define the problem
• This involves developing a researchable question focusing on a specific subject and
figuring out, exactly what it is you want to learn about.
Step 3: Review the existing literature
• This is required in order to learn whether studies have already been done on the subject
and, if so, what the results were.
Goals of a Literature Review
• Demonstrate a familiarity
• Show the path of prior research
• Integrate and summarize
• Learn from others !! Reviewing the literature helps the researcher narrow down the subject, come up with ideas
about specific questions to ask and/or particular areas to explore, and to assess if there is
anything new to learn about the subject that has not already been uncovered by previous studies.
Step 4: Formulating a Hypothesis
• Ahypothesis is a statement you expect to find according to predictions form a theory.
Ex. From Durkheim: suicide rates are dependent upon degrees of social integration
• Hypothesis – an empirically testable statement
• Three-part sequence:
1. Conceptualization – the process of identifying and defining the concepts your
research will address.
• Refining an idea by giving it a very clear, explicit definition.
• Hypothesis require that concepts be formulated.
• Concepts: mental constructs that detail the phenomena to be studied. i.e.
Durkheim’s concept of social integration
• Avariable is a concept whose value may change from one case to another
• It is the unit of analysis to be measured and studied ie. group affiliation
(catholic, male, single, etc), rates of suicide
• Before we set out to measure a variable we must further define it
2. Operationalization – the process of linking a conceptual definition with specific
• We need operational definitions – that is, we need to operationalize a
variable – specify exactly what one is to measure in assigning a value to a
• They help social scientists continue to move form a broad but somewhat
refine conceptual definition, to a more specific definition that they will use
to form their research question.
Ex. Social class – before we can measure people’s social class we must
first decide what it is we will measure
We could choose: Income level, Years of education, Occupational types
3. Measurement – the process of observing concepts, as indicated by their
operational definitions, and assigning some type of score or meaning to people’s
• Variable – a feature of a case or unit that represents multiple types, values
o Independent Variable – the variable of factors, forces, or
conditions acting on another variable to produce an effect or
change in it. (Causes a change in Dependent Variable)
o Dependent Variable – the variable influenced by and changes as
an outcome another variable.
o Intervening Variable – a variable that comes between the
independent and dependent variable in a causal relationship.
(Spuriousness, changes the outcome of the Dependent Variable) Ex. Lower density (more space) reduce crime rate. BUT the
intervening income is INCOME!
Step 5: Choose a Research Method
• A RESEARCH METHODS IS AN ACCEPTED MEANS BY WHICH
YOU COLLECTYOUR DATA
Step 6: Collecting the Data
a. We have to make certain that the operational definitions must measure what they are
intended to measure and not something else.
• Validity: the measurements are consistent with what was intended to be
b. We have to make certain that our operational definitions would illicit the same findings if
the study was conducted again by you or another researcher.
• Reliability: the quality of consistency in measurement. The process must
yield the same results if repeated after time.
Six Types of Research Methods
• The most popular method among researchers
• Involves direct questioning of research subjects
• Practical issues involve: Selecting a sample
o Population: the target group that you intend to study
o Sample: a set number of individuals from among the target
• How one chooses a sample is crucial – the choice may skew results
• If you wish to generalize your findings (generalizability), then your
sample must be representative of your population
o Random sample: every element of the population has the same
chance of ending up in the sample. Important if goal of study is to
o Stratified random sample: a sample for selected subgroups of the
target population in which everyone in those subgroups has an
equal chance of being included in the research
o Snowball sampling: sample is created by a continuous series of
introductions to friends and colleagues. Not very representative or
generalizable. *For hard to reach populations
• Neutral questions: Your questionnaire must be designed to allow
respondents to best express their own opinions. Questions should not be
leading or use biased language.
Ex. Do you think crackheads should be allowed to receive welfare?
Do you think president Obama is doing a good job dealing with economy? Excellent, Very Good, Good, Fair, Poor (This gives them option to define)
StronglyAgree,Agree, Neutral, Disagree, StronglyAgree
o This creates a Response Bias because someone could say