COMM 88 Lecture Notes - Lecture 14: Social Desirability Bias, Internal Validity, Experiment
5/22/18
●Experimental research (continued)
○Threats related to pre-testing (or measures over time):
■Maturation
■testing/sensitization
■Statistical regression (to the mean)
■Mortality
●People that were in the first time of the study, die, drop out, cannot be in
study.
●This is an explanation for a change in results
○Have to check who is dropping out
○Threats (continued)
■Reactivity effects:
●Participants reaction to being studied, rather than to IV/treatment,
influences DV
○Demand
■Characteristics
■participants figured out your hypothesis and are trying to
give you this answer
■This relates to social desirability effect yet social
desirability is broader, the reactivity effect is a result.
○Hawthorne effect
■Productivity keeps going up because people know they
are being watched
■Attention = evaluation = better performance
○Placebo effect
■You react based on getting something and you think you
should act a certain way
■Getting treatment that should have a particular outcome,
so you act this way even though you weren't actually
given the treatment
■Belief drives participants
Document Summary
Threats related to pre-testing (or measures over time): People that were in the first time of the study, die, drop out, cannot be in study. This is an explanation for a change in results. Have to check who is dropping out. Participants reaction to being studied , rather than to iv/treatment, influences dv. Participants figured out your hypothesis and are trying to give you this answer. This relates to social desirability effect yet social desirability is broader, the reactivity effect is a result. Productivity keeps going up because people know they are being watched. You react based on getting something and you think you should act a certain way. Getting treatment that should have a particular outcome, so you act this way even though you weren"t actually given the treatment. All group(s) get equal time, attention, etc. Experimenter"s behavior or attributes, rather than treatment (iv), influences dv. Same thing again (true experiment, etc. ) but also: