The Designs
This category of designs provide almost no control over threats to internal validity
What are pre-experimental designs?
This threat refers to the ability to generalize the results from your study to a larger population
What is selection and X?
This paper concluded that a "simple educational module" produced a 50% improvement in the primary outcome and saved the organization up to 4 million dollars
What is the Zack (ventilator associated pneumonia evaluation) paper?
Surgeons at USC enrolled in a multi-site research study decide to implement their own quality improvement after finding out that they were assigned to the control condition (and UCLA to the intervention)
What is compensatory rivalry?
0 0 0 X 0 0 0
What is an interrupted time series design?
The value of this design is conferred by the multiple pre-tests it includes
What is a time series design?
This threat refers to the influence of being observed or tracked on participant behavior
What is the Hawthorn effect (Reactive Arrangements)?
This reading illustrated the potential bias introduced by simple imputation with differential drop out between the intervention and control condition
What is the Schwartz (tobacco Quitline evaluation) paper?
Older adults participating in a one-group fall prevention study show increases in falls over the study period due to aging
What is maturation?
R O X O O
R O O O
What is a pre-test post-test control group design (true experiment) with two follow-up assessments?
The biggest threat to this quasi-experimental design is selection
What is a nonequivalent control group design?
HPV vaccine increases over time, effect of COVID-19 on inappropriate days of inpatient care
What is a secular trend? What is a history threat?
This paper presented the results of a pragmatic evaluation but saved the good stuff (evidence of internal validity threat) for the discussion
What is the Dowding et al paper (effect of EHR on nursing sensitive outcomes)?
Due to COVID-19, researchers decide to measure participant weight (primary outcome) at follow-up by allowing patients to weigh themselves at home, unlike at baseline when all measures were taken in the clinic by trained personnel
What is instrumentation?
X O
What is a one-shot case study?
This design controls for pre-test sensitization, but doesn't allow investigators to have a proper Table 1
What is a post-test only control group design?
Concluding the intervention had no effect when in reality your study only had 40% power to detect a significant effect
What is low statistical power (statistical conclusion validity threat)?
This paper underscores the importance of carefully understanding both the numerator and denominator when the primary outcome is a rate
What is the Simons-Morton (teenage risky driving intervention) paper?
Your helpful research nurse decides to spend extra time assisting the participant she just randomized to the control group
What is compensatory equalization?
X 0
0
What is a static-group comparison?
The equivalence of this study design to a true experiment depends heavily on how many units are randomized
What is a cluster randomized trial?
This threat may occur when pre-existing differences such as mean age between the intervention and control group leads to differential outcomes at follow-up that are not due to the intervention itself
What is interaction of selection and maturation?
These two papers highlight the potential for politicization of evaluations of high profile public programs
What are the Kotelchuck (WIC) and Baicker (Medicaid) papers?
In the context of a genetic testing promotion study, clinic A participants (Brad Pitt fans) react differently than clinic B participants (Angelina Jolie fans) to Angelina Jolie's announcement about her positive gene test and prophylactic bilateral mastectomy
What is interaction of selection and history?
R 0 (X)
R X 0
What is a separate-sample pretest-posttest design?