The Carter Center
Statement about statistical
assessment of the referendum results
September 5th, 2004.-
Upon receiving complains from the opposition Coordinadora
Democrática about suspected irregularities in the voting
results recorded in the machine, specifically, a pattern of
tied results among 2 or 3 machines in the same voting
tables, The Carter Center consulted with several
statisticians. The statisticians advising our mission and
the OAS mission, an independent statistician from Stanford
University and an independent computer scientist at John
Hopkins University all told us that the actual results are
within the predicted range and do not of themselves prove
fraud. As we had noted, these tied results affect both the
YES and the NO vote.
Professor Jonathan Taylor of Stanford University provided an
analysis to The Carter Center the week of August 16th,
cited with his permission in an article published by
Jennifer McCoy in The Economist on September 2nd.
Subsequently Prof. Taylor found a mistake in one of the
models of his analysis which lowered the predicted number of
tied machines, but which still found the actual result to
lie within statistical possibility. Prof. Aviel Rubin of
Johns Hopkins University says on his webpage “…our
analysis of the same data, based on simulations, did not
detect any statistical anomalies that will indicate obvious
fraud in the election.” (http:/venezuela-referendum.com/)
The Carter Center continues to be open to receiving any
evidence of irregularities or fraud, and we will continue to
inform the Venezuelan public of our assessment. A report on
the entire recall process from November to August will be
forthcoming.
Kay Torrance, The Carter Center,
404-420-5129,
[email protected]
Pedro Antonuccio, The Carter Center,
58-212-9910087,
[email protected]
####
|
|
|