The Basics of Experimentation II: Final Considerations, Unanticipated Influences, and Cross-Cultural Issues
Lecture, Chapter 7
•Samples of participants chosen for an experimental study depend in part on precedent, availability, and nature of the problem.
•If there is an established pattern, or precedent, of using participants from a particular population in literature, we can increase the likelihood of success but limit generalizability.
•Some populations, such as college students, are used for their availability, but again, the limits to this should be considered.
•Often, the type of research project will intuitively suggest the appropriate population.
•The number of participants used in an experiment can be affected by finances, time, and availability of participants.
•Generally speaking, the lower the within-group variability, the fewer participants needed; the inverse is also true.
•The greater the number of participants, the greater the impact around the center of the distribution and the lesser the influence of extreme scores.
•The greater the number of participants, the higher the power (the likelihood of statistical significance) of the results.
•The nature of the IV often depends on the type of equipment needed, but many creative, inexpensive ways of presenting the IV may be used.
•While recording the DV, it is primarily important that the experimenter be as inconspicuous as possible.
•Keep the equipment as simple as possible, understanding that dependence on the equipment can itself be a liability.
Experimenter as an Extraneous Variable
•Experimenter characteristics, such as friendliness (previously found to be predicted by gender), anxiety, or introversion/extroversion may influence results even if constancy has been obtained.
•Experimental characteristics can be controlled through:
Standardized appearance, attitude, etc.
•Experimenter expectancies, as shown in numerous studies, influences experimenter treatment of participants, in turn influencing participant performance, a phenomenon known as the Rosenthal effect.
•Experimenter expectancies can be controlled through:
Preparation of instructions carefully prepared so exact wording and tone used is constant.
Use of automated equipment in instructions or test administration
Use of single-blind experiment, in which the experimenter is unaware of which treatment each participant is receiving.
Participant Perceptions as Extraneous Variables
•Demand characteristics, features of an experiment that inadvertently lead participants to respond in a particular manner, can operate as a confounding or nuisance variable.
•Good participant effect is the tendency of participants to behave as they perceive the experimenter wants them to behave.
•Response bias occurs:
when participants respond either positively (yea-sayers) or negatively (nay-sayers) to all questions, even when doing so involves contradiction.
When the experimental context or environment influences participants to respond in a certain manner or response set.
•Demand characteristics may be controlled:
through both single-blind (participants unaware of treatment used) and double-blind (both
experimenter and participants are unaware).
by giving participants incorrect information about the nature of the project (difficult to get IRB
approval and may still yield erroneous data through incorrect guesses).
by reversing the response required to measure the same attribute.
through close scrutiny of questions and possible pilot testing.
Research and Culture
•Cross-cultural psychology is a branch of psychology whose goal is to determine the universality of research results.
•Some results may be culture-specific, attributable to lasting values, attitudes, and behaviors shared by a cultural group.
•Unfortunately, researchers can have an ethnocentric perspective, in which other cultures are viewed as an extension of their own.
Fundamental Attribution Error
•The tendency to explain behavior in terms of internal dispositions, such as personality traits, abilities, motives, etc., as opposed to external situational factors, is the fundamental attribution error.
•Culture can have an influence on the research question asked, the hypothesis, and the data collection technique.
•Sampling procedure becomes of primary importance, particularly regarding issues of representation and generalizability.
•Researchers must take into consideration the tendency of a particular culture to respond in a certain manner, or the cultural response set.