BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Institute of Statistical Research and Training - ECPv6.15.17//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Institute of Statistical Research and Training
X-ORIGINAL-URL:https://isrt.ac.bd
X-WR-CALDESC:Events for Institute of Statistical Research and Training
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:UTC
BEGIN:STANDARD
TZOFFSETFROM:+0000
TZOFFSETTO:+0000
TZNAME:UTC
DTSTART:20250101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=UTC:20260223T130000
DTEND;TZID=UTC:20260223T143000
DTSTAMP:20260302T142843
CREATED:20260221T064222Z
LAST-MODIFIED:20260222T052903Z
UID:8957-1771851600-1771857000@isrt.ac.bd
SUMMARY:Applied Statistics and Data Science Seminar on Monday 23 February 2026
DESCRIPTION:Title: Win Ratio Estimators for Prioritized Composite Outcomes from Observational Studies and Randomized Experiments with Noncompliance \nVenue\, date and time: ISRT\, 23 February 2026\, 1:45 pm \nSpeaker: Md. Muhitul Alam\, Lecturer\, ISRT\, Deltin 7 Aviator গেম টাকা ইনকাম of Deltin 7 bangladesh \nAbstract: \nWin ratio statistic is a comparative summary of ordered competing risk data based on the number of wins (favourable outcome) among pairs of treated and control subjects. However\, it can yield misleading results in the presence of confounding\, a common issue in observational studies that may also occur in randomized controlled trials (RCTs) when noncompliance is present. If potential confounders are measured\, the recently developed inverse probability weighted (IPW) win ratio offers improvements\, but it is not robust against extreme propensity scores. This study introduces necessary counterfactuals to define a win ratio estimand and proposes several propensity score-based win ratio estimators\, including the stratified-propensity-score-weighted (SPW) win ratio\, stratified-propensity-score-unweighted (SPU) win ratio\, covariate-adjusted (CA) win ratio\, and the overlap-weighted (OW) win ratio. Two variants of the SPU and SPW win ratios are also presented: one is based on quantiles of the estimated propensity scores\, and the other is based on a data-driven algorithm. Simulation results demonstrate that under ideal conditions\, all proposed methods perform comparably to the existing ones; however\, in the presence of extreme propensity scores\, the SPW\, OW\, and CA win ratios show reduced bias and coverage probabilities closer to the nominal level. The study also demonstrates that weighting based on the absolute standardized mean difference (ASMD) can be advantageous compared to SPU win ratio estimators when extreme propensity scores are present\, although it tends to be slightly conservative. Furthermore\, the study reveals that the proposed methods are robust to the misspecification in the propensity score model. This study also introduces win ratio estimators for RCTs in the presence of noncompliance\, where confounders are typically unmeasured. We propose the intention-to-treat (ITT)\, as-treated (AT)\, per-protocol (PP)\, and instrumental variable (IV) win ratios. Extensive simulations demonstrate that the IV win ratio provides unbiased estimates and achieves coverage probabilities close to the nominal level\, even with high rates of noncompliance. Finally\, we illustrate the application of our proposed methods using data from the German Breast Cancer Study and the Job Search Intervention Study (JOBS II).
URL:https://isrt.ac.bd/event/applied-statistics-and-data-science-seminar-on-monday-23-february-2026/
CATEGORIES:seminar
END:VEVENT
END:VCALENDAR