Missing data have been a serious problem in alcohol prevention data bases. Two kinds of missing data, which we call omissions and attrition are present in virtually all longitudinal panel studies. Recent developments in missing value imputation (replacement) and analysis with missing data offer advancements in data analysis previously unavailable in alcohol (prevention) research. Unfortunately, these new methods are available at present to only a few, highly quantitative researchers.
Our aim i s to increase the usefulness of alcohol prevention databases by developing missing value imputation and analysis procedures that apply specifically to problems common in our field. We will solve problems of omission and attrition using the latest methodology. For attrition, our analyses will allow us to draw valid conclusions regarding the effectiveness of several recent alcohol prevention programs as if there were no attrition, thereby addressing recent criticisms of alcohol prevention findings. Finally, we will produce a Technical Assistance Report. This Report will be of value to prevention researchers with typical quantitative backgrounds. The Report will be written in readable form, and will cover problems covered by most prevention and other alcohol research. Finally, the Technical Assistance Report will include detailed examples, including details of the computer software used. Software will also be made available to interested researchers.
| Graham, J W; Donaldson, S I (1993) Evaluating interventions with differential attrition: the importance of nonresponse mechanisms and use of follow-up data. J Appl Psychol 78:119-28 |
