Research Designsx

Experience Carefully
Experimental Research
 Laboratory
experiments allow researchers to
see how things turn out when one and only
one thing is changed
 Manipulation
 Random
assignment (vs matching)
When an additional variable (nuisance
variable) exists that may influence the
dependent variable and that varies
systematically along with the independent
Confounds threaten internal validity. It can be
reduced through random assignment and by
holding conditions constant.
A variable that is held constant in a study but
which influences the relation between the
independent and dependent variables.
Artifacts threaten external validity. It can be
reduced through random selection and
maximizing experimental realism.
Extraneous variable in an experiment that
influences the dependent variable but that is
evenly distributed across the experimental
Noise does not threaten validity, but it
decreases the ability to detect an effect
statistically. It can be reduced through the use
of homogenous samples.
Person Confound
people who are high or low on this variable
also happen to be high or low on some
individual difference variable that is
associated with the outcome variable of
Reduced or eliminated through random
Procedural Confound
Researcher mistakenly allows a second
variable to vary along with a manipulated
Eliminated by repeating the study while
controlling for this variable
Operational Confound- when a measure
designed to assess a specific construct
inadvertently measures something else as
well. It can be eliminated by refining the
operational definition (measure).
 Eliminate
individual differences (person
 Eliminate other kinds of confounds
(procedural and operational confounds)
 Pull researchers into the laboratory
(controlled environment)
 Allow
researchers to observe the invisible
(e.g. fMRI)
 Provide information about interactions
(through stats)
 Minimize noise (reduced variability)
 Problem:
 Solution: Two forms of realism (mundane and
Mundane Realism- the degree to which the
physical setting in an experiment is similar
to the real-world setting in which the
experimenter’s independent and
dependent variables are most likely to
Experimental Realism- the degree to which
the subjective experiences of research
participants are realistic or psychologically
meaningful. Well-designed experiments
can be high in experimental realism even
when they bear little physical resemblance
to the real world.
recipe for experimental realism?
Manipulation checks throughout an experiment
asking research participants about their
 There
is a perceived conflict between
internal and external validity because
confounds threaten internal validity and
artifacts threaten external validity.
highly controlled experimental design
will eliminate confounds, but can also be
considered an artifact. On the other
hand, a loosely controlled experiment is
often subject to many confounding
 Experiments that are high in either
mundane realism or experimental
realism tend to be high in both internal
and external validity.
 Experimental
versus nonexperimental
 Laboratory versus field experiments
 Setting
the stage
 Rehearsing and playing the part:
Be suave
Be nice
Be educated
 Replicate
as needed
- Be honest
- Be a good liar
- Be attentive

similar documents