Analysis approaches to the evaluation of community interventions must
be sensitive to a wide variety of analytic contaminants that may bias
the statistical assessment of changes in outcome measures. These conta
minants include model misspecifications related to failures to control
for community-specific time trends, temporal autocorrelated errors in
equations spatial autocorrelated errors among geographic units, and o
ther failures of unit independence otherwise indexed by estimated intr
aclass correlations. Although an enormous amount of progress has been
made toward the solution of many of these analytic problems over the p
ast years, the contemporary evaluator of community interventions is le
ft with a number of unenviable design and analysis choices; choices th
at inevitably force an assessment of the relative threats of different
sources of error to the internal and external validity of the evaluat
ion. This article describes the choices made for the evaluation of the
Community Trial Project outcome data.