Types of Research Questions
What is the proportion of adults, aged 18-39 years across race/ethnic groups in the US, who are being treated for hypertension in 2023?
Descriptive
What is the framework called for estimating a causal effect?
Potential outcomes or counterfactual outcomes
This causal identification condition describes when there is statistical independence between the potential outcomes and the treatment/action under study, or when treatment status is randomized, which group of participants received the treatment is irrelevant.
Exchangeability
This identifiability condition notes the probability of receiving every level of treatment is positive for every individual. This means that there is no individual for whom receiving the treatment is impossible (and vice versa for the control)
positivity
In adults aged 18-39 years who are being treated for depression, what is the effect of adding on an outdoor walking intervention (30 minutes a day) to standard care on depressive symptoms after 6 months?
Causal
This identifiability condition says that among people who were exposed, the effect of the exposure is no different than it would have been if they had been assigned that exposure/treatment/intervention (same goes for being unexposed/untreated/no intervention)
Exchangeability
This approach to randomization describes when an investigator uses several randomization probabilities that depend (are conditional) on the values of prognostic factor(s) (L)
Conditional randomization
When looking to estimate a causal effect from observational data, this identifiability condition can be best described as “a well defined intervention”.
consistency
What is the association between higher levels of vegetable intake compared to little/no vegetable intake with risk for colon cancer?
Association
“Blank” is defined by a different risk in two disjoint subsets of the population determined by the individuals’ actual treatment value (A = 1 or A = 0), whereas causation is defined by a different risk in the same population under two different treatment values (a = 1 or a = 0)
Association
This analytic approach describes effect is the effect of treatment assignment (or allocation)
Intent to treat
This framework describes an approach to formulating a causal question from observational data?
target trial framework
What sociodemographic factors most strongly predict being screened for type 2 diabetes in adults in Kaiser Permanente Health?
Predictive
What is the fundamental challenge of estimating a causal effect?
Always missing half the data
Investigators undertake this analysis when the goal is to estimate the causal effect that would have been observed if all patients had adhered to the protocol of the RCT
per protocol
Final jeopardy: An analytic approach that provides the best framework for a valid estimate in observational data aligns what following 3 elements?
Eligibility, Time zero, treatment/exposure assignment (non prevalent or non delayed)