A controlled experimental design applied in a field setting was used to determine the effectiveness of a bipartisan mailed letter reminding registered low income black voters to participate in the 1988 Presidential election. Each member of three groups of approximately 85 voters received either one, two or three such reminder letters shortly before the election. A fourth, control group of voters did not receive any letters. Statistical analysis revealed that the reminder letters appeared to have no effect on voting behavior.
Canady, William Kelly and Thyer, Bruce A.
"Promoting Voting Behavior Among Low Income Black Voters Using Reminder Letters: An Experimental Investigation,"
The Journal of Sociology & Social Welfare: Vol. 17
, Article 8.
Available at: http://scholarworks.wmich.edu/jssw/vol17/iss4/8