2. Generality across response measures: Many variables can be defined and
measured in different ways. The variable fear, for example, can be
defined in terms of physiological measures (for example, heart rate), selfreport
measures, or behavior. In a research study, a researcher typically
selects one definition and one measurement procedure. In this case, the
results of the study may be limited to that specific measurement and may
not generalize to other definitions or other measures. For example, a
study may find that a particular therapy is effective in treating phobias
when fear is defined and measured by heart rate. In actual practice,
however, the therapy may not have any effect on phobic patients’
behaviors.
3. Time of measurement: In a research study, the scores for individuals are
measured at a specific time after (or during) the treatment. However, the
actual effect of the treatment may decrease or increase with time. For
example, a stop-smoking program may appear to be very successful if
the participants are measured immediately after the program, but may
have a much lower rate of success if participants are measured 6 months
later. Thus, the results obtained in a research study in which responses
are measured at a specific time may differ from the results obtained
when measured at a different time.