BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Rethinking Weak Instruments with a focus on the Multiple Risk Fact
 ors Setting  - James Lane (University of Cambridge)
DTSTART:20241025T143000Z
DTEND:20241025T160000Z
UID:TALK223033@talks.cam.ac.uk
CONTACT:Martina Scauda
DESCRIPTION:Weak Instrument Bias is a well known phenomenon that occurs wh
 en the association of instrumental variables with the risk factor is small
 er. In other words\, when the relevance assumption is close to being viola
 ted. However\, there are several main deficiencies in the literature on th
 is topic. First\, the failure to account for the increased variance in the
  estimator when weak instruments are present\, not only the bias weak inst
 ruments induce. Second\, how to handle weak instruments when multiple risk
  factors (the multivariable setting) are being used. In this case\, instru
 ments can be unconditionally weak and/or conditionally weak with a given r
 isk factor (conditional on the association with the remaining risk factors
 ). Third\, we identified that the Anderson-Rubin test doubles as a type of
  over identification test which often can lead to an empty confidence set.
  This talk will discuss both of these gaps in the literature and our work 
 so far to address them. 
LOCATION:MR12\, Centre for Mathematical Sciences\, Wilberforce Road\, Camb
 ridge
END:VEVENT
END:VCALENDAR