cansin-r / enosfowler-reanalysis

Experimental Design : ReAnalysis of Enos and Fowler's Stratified Randomized Experiment

SRTReanalysis

Final Project for Experimental Design (Stat 158) at UC Berkeley -- reanalysis of Enos and Fowler's Stratified Randomized Experiment published on March 11, 2014 in an article titled "Pivotality and Turnout: Evidence from a Field Experiment in the Aftermath of Tied Elections". Enos and Fowler leveraged the tied results of the 2010 November Election for Massachussetts State House in the 6th Worcester District in order to study the effects of pivotality in the special election. The existing literature claimed that pivotality, voters thinking that they have a say in whose elected, increases voter turnout. Enos and Fowler's results contradicted the existing literature.

In the first part of the project, I am performing a replication of the original analysis.

In the second part of the project, I am reanalyzing their data using regression readjustment, post stratification and matched pair experiment.

My reanalysis confirms their results.