Covariates: selection, target moments, measurement

14 open problems

Choosing effect modifiers, reconstructing a joint distribution from published marginals, and the fact that the “same” variable is often not the same variable.

ID Problem Priority Verdict Maturity prank
COV-01 Deciding which covariates are true effect modifiers Very high Partially addressed Promising 1
COV-07 Unknown target joint covariate distribution Very high Partially addressed Promising 1
COV-02 No evidence on how the modifier set changes across effect scales High Partially addressed Emerging 2
COV-03 Design-dependent roles of prognostic covariates High Overstated Established 2
COV-04 Selection uncertainty in PAIC interaction shrinkage is uncalibrated High Partially addressed Promising 2
COV-05 Legitimate versus dangerous roles for flexible learners High Partially addressed Promising 2
COV-08 Entropy balancing has no network extension or variance theory High Overstated Promising 2
COV-09 Measurement non-equivalence of covariates across trials High Partially addressed Promising 2
COV-10 Effect-modifier selection does not check interaction consistency High Confirmed open Promising 2
COV-11 Landmark external controls lack covariate inputs for population adjustment High Confirmed open Emerging 2
COV-12 No aggregate integration method for median-based covariate summaries High Confirmed open Promising 2
COV-13 No review method for assembling transferable effect-modifier sets High Confirmed open Emerging 2
COV-14 Per-comparator adjustment sets yield non-comparable PAIC contrasts High Confirmed open Emerging 2
COV-06 No causal forest aggregates to a target known only by marginals Medium-high Partially addressed Emerging 3
No matching items

Back to the full catalog