BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Aurekai//compiled from published calendars//EN
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:events-calendar-utdallas-edu-47ddce57f6ff82bae0ca03f8@aurekai.shop
DTSTAMP:20260918T082100Z
DTSTART:20260918T160000Z
DTEND:20260918T170000Z
SUMMARY:Data Science and Statistics Seminar by Dr. Xu Shi (University of Michigan)
LOCATION:Engineering and Computer Science South (ECSS)\, 2.203
GEO:32.98610;-96.75058
URL:https://calendar.utdallas.edu/event/data-science-and-statistics-seminar-by-dr-xu-shi-university-of-michigan
DESCRIPTION:Title: Addressing Unmeasured Confounding in Observational Studies: Advances with Negative Control Methods and Proximal Causal Inference Abstract: Unmeasured confounding remains one of the most significant threats to the credibility of findings from observational studies. Recent developments in negative control methods\, also known as proximal causal inference\, offer promising strategies to strengthen causal conclusions. These approaches leverage negative controls —variables that have no direct causal relationship with either the exposure or the outcome — to detect and adjust for unmeasured confounding. In this talk\, I will review the foundations of negative control methods and introduce the double negative control framework. I will then present our recent work extending this framework to settings where some candidate negative control variables may themselves be invalid. I will conclude with a discussion of open challenges and future research directions.\n\nCompiled by Aurekai from what utdallas.edu published. Times and details are theirs; check the source before you rely on them.\nhttps://calendar.utdallas.edu/event/data-science-and-statistics-seminar-by-dr-xu-shi-university-of-michigan
END:VEVENT
END:VCALENDAR