Rethinking Weak Instruments with a focus on the Multiple Risk Factors Setting

Add to your list(s) Download to your calendar using vCal

• James Lane (University of Cambridge)
• Friday 25 October 2024, 15:30-17:00
• MR12, Centre for Mathematical Sciences, Wilberforce Road, Cambridge.

If you have a question about this talk, please contact Martina Scauda.

Weak Instrument Bias is a well known phenomenon that occurs when the association of instrumental variables with the risk factor is smaller. In other words, when the relevance assumption is close to being violated. However, there are several main deficiencies in the literature on this topic. First, the failure to account for the increased variance in the estimator when weak instruments are present, not only the bias weak instruments induce. Second, how to handle weak instruments when multiple risk factors (the multivariable setting) are being used. In this case, instruments can be unconditionally weak and/or conditionally weak with a given risk 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.

This talk is part of the Causal Inference Reading Group series.

This talk is included in these lists:

• All CMS events
• All Talks (aka the CURE list)
• CMS Events
• Causal Inference Reading Group
• DPMMS Lists
• DPMMS info aggregator
• DPMMS lists
• Hanchen DaDaDash
• Interested Talks
• MR12, Centre for Mathematical Sciences, Wilberforce Road, Cambridge
• School of Physical Sciences
• Statistical Laboratory info aggregator
• bld31

Note that ex-directory lists are not shown.