Please enable JavaScript.
Coggle requires JavaScript to display documents.
Threats to validity - Experiments - Coggle Diagram
Threats to validity - Experiments
Internal validity
problems about drawing inferences about covariation if treatment and variable have causal relationship
Participants
Regression
will naturally do better on postest
Selection
select 'better' participants
Maturation
ppl change/develop
Mortality
dropouts - prob comparing grps
History
events btw pre & postest that influence outcome
Interaction with selection
Treatment
compensatory rivalry
btw grp rivalry
Resentful demoralization
control grp demoralized
Compensatory equalization
inequality of benefit of intervention
Diffusion of treatment
control grp learns of experiment
Procedure
Instrumentation
change of instrument
administered @ diff times/ways
Testing
familiar with measures/responses
External validity
problems drawing inference to other treatment variables, ppl, settings, measurements
Interaction of selection and treatment
Inability to generalize beyond test grp
Interaction of setting and treatment
Inability to generalize to other settings
Interaction of history and treatment
Prob generalizing to past & future situations
Construct validity
validity of constructs/variables
Statistical conclusion validity
appropriate use of statistics