Research
Methods in Psychology
Descriptive Methods
Naturalistic
observation
Intensive
individual case study
Surveys/questionnaires/interviews
Correlational
studies
The Experimental Method
for determining cause-effect relationships
Naturalistic
Observations by Jane Goodall
Observing
Teens at the Mall
Observing
Kids at Play
Using
Various Research Methods
to Study a Behavior Problem
Like ADHD
Attention
Deficit/Hyperactivity Disorder (ADHD)
Diagnosed when a child shows
6 or more
symptoms of inattention
also 6 or
more symptoms of hyperactivity,
These
symptoms are present at an early age & must have been present for at least
6 months.
Naturalistic
Observation
Systematic, unobtrusive observations in classrooms show that kids with ADHD:
cant stay in their seats or sit still, dont pay attention,
dont complete work, are impulsive, are rude to others, get into trouble more,
& lose their temper more often.
This method makes use of real-life situations, but it is
important to use well-trained, unbiased observers.
Survey
Results:
5x more males than females
3-5% of USA schoolkids are taking Ritalin for ADHD
50-60% of ADHD kids show defiant behavior & higher risk
of conduct problems as teens
30-60% continue to have symptoms as adults
40% have a parent with symptoms
Correlation
Correlation: the degree to which one variable or set of data
is related to another variable/set of data.
Correlation coefficient: number between -1 and +1 showing the strength and direction
of this relationship.
Correlations help us predict behavior but do not indicate
the cause of the relationship.
Remember: Correlation does not prove causation.
Twin Studies
Look at Correlations
Is
Hyperactivity Correlated in Twins?
Pairs of male fraternal twins show almost no
correlation (+.05) in their level of motor activity
But pairs of male identical twins show a strong
correlation (+.71) in their level of motor activity
Supports the hypothesis that genetics play a role in ADHD
but does not prove a cause-effect relationship.
Pros/Cons
of Other Methods
Survey: Easy to collect lots of data but may be biased if
sample is poor or responses are not accurate
Case studies: Can provide in-depth data on an individual but
we cant assume it will apply to all others
Experiment: Most able to identify cause-effect relationships
but sometimes results dont generalize to real-life situations
Why
are experiments different?
Compare the behavior of 2 or more groups of participants
under very controlled conditions.
Groups are treated as similarly as possible EXCEPT for the
critical variable(s) (the independent variable) that the researcher is
interested in. The researcher intentionally manipulates or varies the
independent variable to study its impact on behavior.
Random Assignment
To make the 2 groups as
equal as possible, most often participants will be randomly assigned to 1 group or the other. This assures
that there are no systematic differences between the groups.
Why
are experiments different?
If everything is kept constant except for the independent
variable, then any differences in performance between groups should be caused
by that independent variable.
In other words, the experiment tests whether the independent
variable causes changes behavior.
Definitions:
Experimental group - the group of participants exposed to
the independent variable that the researcher is really interested in
Control group - the group not exposed to the independent
variable of interest but rather some substitute control condition.
Definitions:
Independent variable- what the researcher manipulates or
varies; the thing that is different in the experimental group versus the control group.
Dependent variable - the behavior that is observed,
measured, tested; the actual data collected from both groups.
Random
Assignment
Assigning participants to the groups in a study such that
all subjects have an equal chance of being assigned to any group of condition.
Random assignment avoids any systematic differences between the
groups as long as the size of your groups is sufficient
Operational
Definition
A definition that describes the actions or operations that will be made to measure,
manipulate, or control a variable in an
experiment
Quasiexperiments
Sometimes we need to use
pre-existing groups in research (e.g. males vs females, alcoholics vs
non-alcoholics). Since we cant randomly assign participants to groups, there may other differences between the
groups that impair our ability to draw
conclusions.
Hypothesis
A tentative statement or prediction about the relationship
between 2 or more variables
Example:
Consumption
of alcohol will impair short-term memory.
Relaxation
training will reduce test anxiety and improve test performance.
Random
or
Extraneous Variables
Any other variables besides the independent variable that seem likely to influence the dependent variable in a particular study
Every effort must be made to assure that the experimental
and control groups do not differ with respect to these extraneous variables
Confounding
of Variables
When it is difficult to separate the effects of an
extraneous variable from those of the independent variable
Confounding of variables interferes with the ability to draw
conclusions about the effects of the independent variable on the dependent
variable
Avoiding
bias
sampling bias
social-desirability bias in self-report data
experimenter bias
bias due to expectations (placebo effect)
Key
features of the scientific process
always looking for alternative explanations
replication of findings