Research Associate · Statistics & Causal Inference
Jinghao Sun.
University of Pennsylvania
I develop causal inference and randomized trial methods for complex temporal data, with applications in public health, medicine, and policy.
Areas of inquiry
Research themes
Causal inference over time
Methods for continuous-time processes, longitudinal treatments, and panel data.
Clinical trials & survival
Randomized trial methods for survival outcomes and restricted mean survival time.
Robust identification
Partial identification and inference when standard assumptions may not hold.
A starting point
Selected work
On a debiased and semiparametric efficient changes-in-changes estimator
Journal of the Royal Statistical Society: Series B · Major revision
Semiparametric methods for causal inference with changes-in-changes designs in panel data.
Beyond fixed restriction time: Adaptive restricted mean survival time methods in clinical trials
Biometrika · In press
Adaptive restricted mean survival time methods for clinical trials with survival outcomes.
Causal identification for continuous-time stochastic processes
arXiv preprint · Manuscript under revision
Identification of causal effects when treatments and outcomes evolve in continuous time.
Recent updates
News
-
The adaptive restricted mean survival time paper appeared online as an accepted manuscript in Biometrika.
Journal article -
I began a Research Associate appointment at the University of Pennsylvania.
About -
I presented the changes-in-changes work at the Joint Statistical Meetings in Nashville.
Related paper
More news
-
I presented an adaptive restricted mean survival time poster at the American Causal Inference Conference in Detroit.
Related paper -
I received the Best Student Paper Award at the Lifetime Data Science Conference.
Honors
Research tools
Software
R package
AdaRMST
Adaptive restricted mean survival time methods for clinical trials, especially under non-proportional hazards.
R package
crc.partialid
Partial identification analysis for capture-recapture experiments.