Chapter 2: Scientific Research Methods
Why do Scientific Research?
-One of the major goals in comm. Psych is to create or engage in some form of
social change so that individuals and communities can benefit.
-Scientific research is a method that provides psychologists a way to distinguish
effective from less effective changes.
Areas of concern must be identified and described
-Factors related to these problems must be articulated
-Based on this articulation possible interventions/solutions can be created and
-Once a program is found effective it still has to be determined if the intervention
can be successfully implemented in particular community contexts
-If implementation is successful, it has to be determined if the program can be
launched on a broader scale
-If the program is successful, the researcher is left to re-examine the community
status and see if any other needs may exist
What is Scientific Research?
Theory: Systematic attempt to explain observable or measureable events relating
to an issue. Ex; Homelessness or Alcoholism
-The goal of a theory is to allow researchers to describe, predict, and control for
why or how a variable relates to observable or measureable events pertaining to
Model: Working blueprint of how a theory works
Paradigm: A smaller component from within the model that guides researchers to
conceptualize specific event sequences
-Kuhn uses gives paradigm two meanings:
1. To describe a set or collection of ideas, values and theories that are
commonly agreed on in a sociological way to guide the direction and
conduction of the scientific inquiry
2. As a concrete puzzle solution to given problem. Term is used to describe
the progression from theory to model to paradigm
Falsifiability: Popper (1957/1990) – Science is about the continuous test of a
theory. Continuous testing assumes that it is possible for a theory to be false.
Because a theory can be false, it must always be tested.
Example: For years researchers investigating alcoholism conceptualized
excessive drinking as a consequence of genetic predisposition, using the medical
explanation of alcohol as a disease. Yet in recent years researchers have begun
to challenge the genetic disease theory of alcoholism, they argued that
environmental factors such as prolonged unemployment may be a reason for
excessive alcohol use. Thus a new theory emerges – the distress theory of
alcoholism. This new idea leads to a paradigm shift from genetics to environment,
which leads to the development of new models. Scientific Revolutions
-Kuhn argued that major scientific development is not linear
-Cause a dramatic shift in the way we see our world
-Causes changes in our theories, models and paradigms
The Fidelity of Scientific Research
Reliability: Extent to which measureable features of a theory are trustworthy or
-If two observers rate what they saw and they agree its called Inter-
-If the question is asked twice within the same set of questions its called
-If the question is asked twice at two separate occasions across time its
called test-retest reliability.
Internal Validity: The degree to which we believe the results of a study truly
describe what happens in a given set of research conditions
-Research is said to have a high internal validity when confounding variables are
at a minimum.
Confounding Effects: unrelated variables that influence the dependent variable
and invalidate the conclusion drawn from the research.
External Validity: refers to the generalizability of results to the larger world.
-When trying to generalize results to a given population, a representative sample
of that population is taken. Random Sampling: allows for all potential
participants to have an equal chance of being selected for a study
Diffusion of Treatment: A consequence of biased sampling. It is difficult to draw
definitive conclusions about the respective efficiency and effectiveness of a
program because the program isn’t pure. Effects of one treatment have spilled
Traditional Scientific Research Methods
-First step in devising a study is to conduct a review of scientific literature. The
researcher then draws what appears to be a logical conclusion from reading the
-Hypothesis: A statement of what might be expected from the study.
-Researcher chooses b/w two research designs to use in the study to examine
the hypothesis experimental design or correlational design
Design: Systematic plan to test the hypothesis
-These designs are used to guide what kinds of data are gathered from groups of
people and how data gathering is carried out.
Population: The group of people the research is attempting to understand
-Getting data on an entire population is difficult therefore we use samples.
Sample: Subset of the population that is supposed to represent the population
Random Sample: Every member of the population has an equal chance of being
selected Convenience Sample: Members are chosen for no other reason other than they
are available. Example: College students, because they are readily available to
participate in research conducted at universities.
Stratified Sample: Tries to match the known characteristics of the population.
Example: if we know 40% of the population is male, we would try to get a sample
that is 40% male.
Purposive Sample: One chosen for a specific reason. Example: In a test of drug
use among pregnant women, only pregnant women would be chosen.
-Correlational Methods include a class of designs (eg. surveys) and measurement
procedures, as well as techniques (eg. Self report), that allow one to examine the
associations/relationships between two or more variables in their natural
-Causation cannot be determined because other unstudied variables could have
produced the effects noted.
-Associations can be spurious (false while giving the appearance of being
correct) when confounding variables are responsible for relationships
Pearson Correlation Coefficient: A statistic used to quantify the association’s
b/w two variables. Ranges from +1.00 to -1.00.
-Considered the gold standard for research
-Include a class of designs and measurement procedures that allow one to
manipulate independent variables and observe the resulting effects on dependent
Independent Variable: is the condition that is varied between groups
Dependent Variable: What scientists measure to see the effects of the
Pretest-posttest control group design: is a common design which involves
assessing a dependent variable before and after a experimental manipulation in
one group (experimental group) and before and after no manipulation condition in
the control group.
-Assignment to the control group is random
-Since the assignment is random the two groups (experimental and control
groups) can be assumed to be equivalent. Therefore if there are any differences
b/w the group ir can be assumed that the independent variable is what brought
the change because it is the only difference b/w the two groups.
-When an experiment would be unethical to perform a quasi-experimental design
-It approximates experimental conditions and random assignment but it isn’t quite
able to get all the necessary conditions for a true experimental design
-Non-equivalent pretest-posttest control design: Comparing a group before
and after some experimental manipulation or treatment with another group that ha