Cecilia Mo助教授講演会開催のお知らせ

投稿日:2019年03月14日 カテゴリー:イベント情報  投稿者:学会事務局

東京大学社会科学研究所では、UCバークレーとの戦略的パートナーシップ事業の一環として、実験の手法を用いた政治経済行動論、社会政策研究がご専門のCecilia Mo助教授の講演会を開催することとなりました。Mo先生は、近年4つの学会賞を獲得された新進気鋭の若手研究者です。3月25日(月)、3月26日(火)の両日に本郷キャンパスで異なる内容の講演会を行います(講演会の詳細につきましては、以下を御覧下さい)。



講演会 "When Do the Advantaged See the Disadvantages of Others? A 
Quasi-Experimental Study of National Service"
日時:3月25日(月) 17時30分~19時(17時開場)
会場:東京大学本郷キャンパス 赤門総合研究棟549号室
要旨:Are there mechanisms by which the advantaged can see the perspectives 
of the disadvantaged? If advantaged individuals have prolonged 
engagement with disadvantaged populations and confront issues of 
inequality through national service, do they see the world more through 
the lens of the poor? We explore this question by examining Teach For 
America (TFA), as TFA is a prominent national service program that 
integrates top college graduates into low-income communities for two 
years and employs a selection model that allows for causal inference. A 
regression discontinuity approach, utilizing an original survey of over 
32,000 TFA applicants and TFA’s selection data for the 2007–2015 
application cycles, reveals that extended intergroup contact in a 
service context causes advantaged Americans to adopt beliefs that are 
closer to those of disadvantaged Americans. These findings have broad 
implications for our understanding of the impact of intergroup contact 
on perceptions of social justice and prejudice reduction.

講演会 "Observational Open Science: An Application to the Literature on 
Irrelevant Events and Voting Behavior"
会場:東京大学本郷キャンパス 赤門総合研究棟549号室
要旨:Replication and transparency have become increasingly important in 
bolstering the credibility of political science research, yet open 
science tools are typically designed for experiments. For observational 
studies, which make up the bulk of political science, current practice 
suffers from an important pathology: just as researchers can often 
``p-hack’’ their way to initial findings, it is often possible to ``null 
hack’’ a finding away through specification and case search. We propose 
an observational open science (OOS) framework that consists of (1) 
leveraging the passage of time to add new out-of-sample cases, (2) 
independent collection of all data, (3) pre-registration of alternative 
specifications, (4) multiple simultaneous replications, and (5) 
collaboration between original authors and skeptics. We apply the 
approach to three highly cited studies in the literature on 
``irrelevant’’ events and voting behavior. On some dimensions, the 
studies all replicated quite well. On other dimensions, they replicated 
quite poorly. Had we sought to debunk any of these three studies in a 
conventional approach to replication with ex post specification search, 
we could have easily done so. However, given that we conducted 
pre-registered analyses, we have a more accurate sense of the full and 
complicated picture. We conclude with suggestions for future refinements 
to our approach.