<div dir="ltr"><p class="MsoNormal" style="font-size:12.8px"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial">Hi everyone!</span></p><p class="MsoNormal" style="font-size:12.8px"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><br></span></p><p class="MsoNormal" style="font-size:12.8px"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial">Welcome to the Applied Statistics Workshop 2017-2018!  Our first session will be this Wednesday, September 6.</span></p><p class="MsoNormal" style="font-size:12.8px;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif"> </span></p><p class="MsoNormal" style="font-size:12.8px;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif">This week at the<span class="gmail-m_-3131546366916696156gmail-apple-converted-space"> </span></span>Applied Statistics <span style="line-height:14.72px"><font face="Arial, sans-serif"><span style="font-size:10pt">Worksho<wbr>p we will be welcoming</span></font><span class="gmail-m_-3131546366916696156gmail-apple-converted-space" style="font-family:Arial,sans-serif;font-size:10pt"> </span><b style="font-family:Arial,sans-serif;font-size:10pt"><i>Stephen Raudenbush</i></b><font face="Arial, sans-serif"><span style="font-size:10pt">, a Professor of Sociology at the University of Chicago. He will be presenting work entitled</span></font><b style="font-family:Arial,sans-serif;font-size:10pt"><i> </i></b><font face="Arial, sans-serif"><span style="font-size:13.3333px"><b><i>Estimands and Estimators for Multi-Site Randomized Trials</i></b></span><span style="font-size:10pt">.  Please find the abstract below.</span></font></span></p><p class="MsoNormal" style="font-size:12.8px;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif"> </span></p><p class="MsoNormal" style="font-size:12.8px;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif">As usual, we will meet at noon in CGIS Knafel Room 354 and lunch will be provided.  See you all there!</span></p><p class="MsoNormal" style="font-size:12.8px;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif"> </span></p><p class="MsoNormal" style="font-size:12.8px;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif">-- Dana Higgins</span></p><p class="MsoNormal" style="font-size:12.8px;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif"> </span></p><p class="MsoNormal" style="font-size:12.8px"><span style="font-size:10pt;line-height:15.3333px;font-family:Arial,sans-serif"><br><b><span style="background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial">Title:</span></b><span class="gmail-m_-3131546366916696156gmail-apple-converted-space"><span style="background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"> </span></span><u><span style="background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial">Estimands and Estimators for Multi-Side Randomized Trials</span></u><br><br><b><span style="background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial">Abstract:</span></b><span class="gmail-m_-3131546366916696156gmail-apple-converted-space"><span style="background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"> </span></span></span>In a multi-site randomized trial, sites such as schools or hospitals are sampled; within each site, persons are assigned at random to treatments. Such studies are increasingly common in social welfare, medicine, and education. In this talk, I’ll first use potential outcomes and a super-population framework to precisely describe different potential populations and parameters of interest, which may diverge considerably when treatment effects vary. Second, I’ll show that maximizing a weighted two-level likelihood produces consistent estimators of all parameters, but only after we introduce a correction for estimating between-site variance components. Third, we’ll see that these weighted estimators, while consistent, may be embarrassingly inefficient (to the point of being improved by throwing out data). Precision weighting may help but may introduce large-sample bias. In the interest of time, I will focus on two parameters: (1) the average impact of treatment assignment (“intention to treat effect”); (2) in trials with non-compliance, the average impact of participation in the treatment on those induced by random assignment to participate (“complier average causal effect”). I’ll illustrate with data from the National Head Start Impact Study.</p>
</div>