EXTERNAL
VALIDITY
The extent to which
results can generalize beyond the original sample and research setting.
THREATS TO EXTERNAL VALIDITY
v
Reactive testing
v
Demand characteristics
DEMAND CHARACTERISTICS
- Any
aspect of the study that informs participants about the purpose of the study
– specifically, the hypothesis being investigated.
- The
concern is that participants will act to try to confirm, or disconfirm, the
hypothesis.
- Researchers
may therefore try to disguise the purpose of the study, or study
participants in such a way that they are not aware that their behavior is
being measured.
v
Reactivity
REACTIVITY
- Being
aware that someone is observing or measuring your behavior may change the
way you behave.
- Allowing
time for participants to get used to being observed or measured minimizes
reactivity.
- Nonreactive
or unobtrusive measures are indirect measures of behavior - participants are
not aware of the measure.
v
Selection biases (includes volunteer bias)
v
Experimenter bias
- May
be a problem when the experimenter knows the hypothesis and is aware of
which condition each participant is in.
- The
concern is that the experimenter may treat participants differently in the
various conditions in order to confirm the hypothesis. Also, the
experimenter may record or interpret behaviors differently in the various
conditions.
- These
effects are usually unintentional.
- These
problems may be addressed by the use of the following procedures:
- Single
blind: The researcher does not know the hypothesis.
- Double
blind: Both the researcher and the participants are unaware of predicted
outcomes.
v
Multiple treatment interference
Steps taken to increase internal validity may decrease
external validity and vice versa.
Internal validity may be more important in basic
research; external validity, in applied research.