Chapter 15 Flashcards
What is statistical conclusion validity?
Appropriate use of statistical procedures to assess the relationship between the independent and dependent variables
What are potential threats to statistical conclusion validity?
Low statistical power
Violated assumptions of statistical tests
Reliability and variance
Failure to use intention to treat analysis
What is internal validity?
Potential for confounding factors to interfere
with the relationship between independent and
dependent variables
What are internal threats to internal validity?
history
maturation
attrition
testing
instrumentation
regression to the mean
selection
What are social threats to internal validity?
Diffusion or imitation
Compensatory
equalization
Compensatory rivalry
Demoralization
How can you rule out threats to internal validity?
Random assignment and control groups
* May control threats due to history, maturation,
selection, regression, testing, instrumentation,
and selection interactions
Blinding subjects and investigators
* May control threats due to attrition, imitating
treatments, or compensatory reactions
What are threats to construct validity?
Operational definitions
Comprehensive measurements
Subgroup differences
Time frame
Multiple treatment interactions
Experimental bias
What is the hawthorne effect?
One possible type of experimental bias
The effect of subjects’ knowledge that they
are part of a study on their performance
* First described related to a series of experiments
on workers’ performance
* However a number of flaws in the original
experiments
What is external validity?
extent to which results can be generalized
What are threats to external validity?
Influence of selection
‒ Adherence
Influence of settings
‒ Ecological validity
Influence of history
What are strategies to control for subject variability?
Random assignment
Homogeneous samples
Blocking variables
Matching
Repeated measures
Analysis of covariance (ANCOVA)
What is compliance?
Getting the assigned treatment
Being evaluated according to the protocol
Adherence to protocol requirements
How can data go missing?
Dropouts
Missing a test session
Missing outcome measures
How can you analyze results with noncompliance?
Per-protocol analysis
Intention to treat analysis
What is pre-protocol analysis?
Only includes those subjects who complied with
the trial’s protocol
Non-completers are removed from the analysis
May make the experimental group look more successful as compared to the control group
