10 Analysis with Missing Data in Prevention Research

Missing data are pervasive in alcohol and drug abuse prevention evaluation efforts: Researchers administer surveys, and some items are left un-answered. Slow readers often leave large portions incomplete at the end of the survey. Researchers administer the surveys at several points in time, and people fail to show up at one or more waves of measurement. Researchers often design their measures to include a certain amount of " miss-ingness " ; some measures are so expensive (in money or time) that researchers can afford to administer them only to some respondents. Missing data problems have been around for years. Until recently, researchers have fumbled with partial solutions and put up only the weakest counterarguments to the admonitions of the critics of prevention and applied psychological research. Things have changed, however. Statistically sound solutions are now available for virtually every missing data problem,