The document summarizes the results of a field experiment. It shows that observable characteristics were balanced across the treatment and control groups through a balance table. The main estimate found that resumes with African American sounding names received a callback rate that was 3.2 percentage points lower than resumes with white sounding names, whose average callback rate was 9.65%. It also noted that callback rates increased by 3.3 percentage points for every 10 years of experience. The remainder of the experiment results are in an accompanying log file.
The document summarizes the results of a field experiment. It shows that observable characteristics were balanced across the treatment and control groups through a balance table. The main estimate found that resumes with African American sounding names received a callback rate that was 3.2 percentage points lower than resumes with white sounding names, whose average callback rate was 9.65%. It also noted that callback rates increased by 3.3 percentage points for every 10 years of experience. The remainder of the experiment results are in an accompanying log file.
The document summarizes the results of a field experiment. It shows that observable characteristics were balanced across the treatment and control groups through a balance table. The main estimate found that resumes with African American sounding names received a callback rate that was 3.2 percentage points lower than resumes with white sounding names, whose average callback rate was 9.65%. It also noted that callback rates increased by 3.3 percentage points for every 10 years of experience. The remainder of the experiment results are in an accompanying log file.
Field Experiment Application Experiment results have a typical form in showing the results: Balance table: shows that the randomization worked by comparing observable characteristics across the treatment and control groups. It seems to have worked here. Note: the assumption is that the unobservable characteristics are not related to the treatment indicator. If observable characteristics happen to be related to it, then it would be a tougher sell to claim that the unobservables are balanced. The main estimate is that the resumes with names that are common among African Americans have a call back rate that is 3.2 percentage points lower than resumes with names that are common among white applicants whose mean call-back rate is 9.65%. The coefficient on years of experience is 0.0033. This means that the call back rate increases by 3.3 percentage points with 10 years of experience. This is another way to put the main estimate into context. The remainder explored the resume experiment, and the results are in the accompanying .log file.