Introduction
When possible use experiments
When experiments are not possible (e.g., evaluating the effectiveness of a change in state law) consider using quasi-experimental analysis
Campbell’s (1969) article and running example of analyzing the effect of a change in a speeding law on highway safety
Threats to Internal Validity
Running example: alternative explanations for a decrease in fatalities
History
dry weather
Maturation
safety belt use increase
Instability
bus accident
Testing
newspaper reports of problem
Instrumentation
rates per mile driven estimated by gasoline expenditures
Regression artifacts
new laws applied only when problem unusually bad
Selection
comparison states different in important ways
Selection-maturation interactions
Experimental mortality
population decrease
Threats to External Validity
Interaction effects of testing
extensive pretesting sensitizes population
Interaction of selection and experimental treatment
population unusually responsive to treatment
Reactive effects of experimental arrangements
experimental setting unusual
Multiple treatment interference
attempt to generalize to single treatment
Irrelevant responsiveness of measures
measures responsive to unknown effects
Irrelevant replicability of treatments
treatments cannot be replicated
KARTLÄGGNING AV EGEN SITUATION SWOT (STRENGTHS WEAKNESSES OPPORTUNITIES THREATS)
SAMPLE STRENGTHS WEAKNESSES OPPORTUNITIES THREATS (SWOT) EXERCISE STRENGTHS WEAKNESSES
SWOT ANALYSIS YOUR LIBRARY’S STRENGTHS WEAKNESSES OPPORTUNITIES THREATS WORKSHEET
Tags: (1969) ===============, campbell’s (1969), quasiexperimental, validity, campbell, (1969), designs, threats