The document summarizes key concepts and methods in social science research. It outlines three main methods: 1) Observational which aims to describe phenomena, 2) Correlational which predicts relationships between variables, and 3) Experimental which tests causality. Key aspects of each method are defined such as variables, sampling, controls and random assignment in experiments, and threats to internal and external validity. Feedback loops are also introduced as positive reinforcing effects or negative balancing effects.
2. Three Simple Steps to Social Science
(easier said than done)
STEP 1: Select some concepts of interest
(variables)
STEP 2: Posit (suggest) some relationship
between these concepts (Hypothesis)
STEP 3: Test these suggestions empirically to
see if they are right.
3. Research Methods
Table 2.1 Summary of Research Methods
Method Focus Question Answered
1. Observational Description What is the nature of the phenomenon?
2. Correlational Prediction From knowing X, can we predict Y?
3. Experimental Causality Is variable X a cause of variable Y?
4. I. Observational Method
• What is it? Observing people and carefully
recording measurements of their behavior.
• This includes:
– Ethnography: method of attempting to
understand a group from the inside (i.e. from their
point of view)
– Archival Analysis: observing the accumulated
'archives' (documents) of a culture.
5. II. Correlational Analysis
• What is it? Two variables are systematically
measured, and the relationship between them
is assessed.
• Correlation coefficient is a statistic that tells
you the strength and direction of a
relationship.
• Surveys: research in which a representative
sample of people are asked questions about
their attitudes or behavior.
7. Directionality
• Variables that vary in the same
direction have a positive
relationship.
• Variables that vary in the
opposite direction have a
negative relationship.
8. Directionality
• Example: a decrease in
employment is associated with
a decrease in income:
– Even though both variables go
down, they vary in the same
direction. This is a positive
relationship! (-1 * -1 = + 1)
9. Variables:
Independent (X) vs Dependent (Y)
• Independent variable (X) = the cause.
Variable that influences.
• Dependent variable (Y) = the effect. Variable
that is influenced by the cause; it is dependent
on the cause.
• INCA: the INdependent variable is the CAuse.
10. III. Experiment
• An experiment involves manipulating the
independent variable (X) and observing the effect on
the dependent variable (Y)
• Experiments are the only means by which we can
explore causal relationships; only way we can know
for sure if changes to X cause changes in Y.
• Experimenter needs two dependent variable (Y)
groups of Y:
1. Experimental group- receives ‘treatment’ of independent
variable (X)
2. Control group- does not receive treatment; is left alone.
11. III. Experiment
• Imagine a scientist testing the
effect that some drug, X, has on
growth of rats, Y.
• To see how the drug effects rat
growth, the experimenter will
compare growth in two groups
of rats: Y₁ , the group of rats
that gets the drug (X) and a
group of rates Y₂ that will not.
• Y₁ is the experimental group,
and Y₂ is the control group.
12. III. Experiment
• One assumes separation or isolation
between the setting where X is
applied and the control, where X isn’t
applied.
• It is important that rats which receive
the drug and rats which do not be
alike in all relevant characteristics and
conditions, so that any observed
differences between rats which
receive the drug (the experimental
group) and those that do not (the
control group) can be attributed only
to the drug (X), and not to something
else.
13. III. Experiment
• Random Assignment to condition-
is the process whereby all
participants have an equal chance
of taking part in any condition of
the experiment.
• The purpose is to ensure that any
potentially relevant differences
between the experimental and
control groups are distributed
evenly and therefore won’t affect
the outcome (i.e. will cancel each
other out)
14. III. Experiment
• A counter-factual refers to something
that did not happen, but could have or
would have occurred.
• We use the ‘control group’ to make a
counterfactual argument, which says that:
“in the absence of X, this is how Y₁ would
have behaved.” We assume that Y₁ would
have behaved like Y₂, the control.
• Why? Because they are alike in all
relevant characteristics so any difference
we observe must be a result of the
independent variable, X.
15. III. Experiment
5 Rules for Doing True Experiments
1. Have at least two groups (control and experiment)
2. Randomly assign people to groups
3. Treat the experimental group by manipulating the
independent variable
4. Observe the effect of the treatment on the dependent
variable in the experimental group
5. Compare the dependent variable differences (the
outcome of treatment) in the experimental and
control groups
16. Sampling
• A Sample is a portion of the larger population
that you will study to make inferences about
the larger population.
• General rule: the more diverse a population
is, the larger the sample needs to be!
• Samples should be random: every element in
the population has the same probability of
being in the sample.
17. External and Internal Validity
• Internal Validity = making sure that nothing
besides the independent variable can affect
the dependent variable.
– Achieved by using Controls and Random
assignment to conditions.
• External Validity = extent to which the results
of a study can be generalized to other
situations and to other people.
19. Feedback
• Feedback Loop: occurs when changes
generate effects that then influence the
original causes of the change, making
subsequent change either more (+) or less (-) likely.
Effect Cause
20. Feedback
Two types of Feedback: Positive Feedback
1. Positive (reinforcing, amplifying):
Initial changes become amplified or
magnified over time; patterns are Population Births
reinforced. + +
– Examples: exponential population
growth; nuclear explosion; ‘rich getting
richer’, etc.
2. Negative (counteracting, Negative Feedback
balancing):
Initial changes are counteracted or
Force of
balanced out, so that conditions remain Jump up
Gravity
relatively stable. +
-
– Examples: homeostasis; a thermostat;
“what goes up, must come down”, etc.
Editor's Notes
Comparisons are made with the assumption that events in the test condition have not affected events in settings where the test condition is absent.
Comparisons are made with the assumption that events in the test condition have not affected events in settings where the test condition is absent.
A “fact” is something that does exist or did happen. Therefore a counter-fact is something that does not exist or did not actually happen.
Feedback occurs when the output of a system is also an input to that same system so that a change in a condition in one part of the system creates results elsewhere in the system that in turn change the original conditions. Some examples of feedback concepts in the social sciences include vicious circles, self-fulfilling prophecies, homeostatic processes, and invisible hands (Richardson 1991). Feedback implies circular, or reciprocal causal relations, where A influences B, and B in turn influences A, and so on. Note: the term “cybernetic” refers to goal-oriented or purpose driven systems. This includes all living organisms, and many non-living systems such as thermostats and heat-seeking missiles. These systems use negative feedback in the sense that they vary their output (behavior) so that the difference between their sensory inputs (perceptions) and their goals (‘reference standards’) is minimized.