This presentation is intended to encourage evaluators to consider the use of AutoRegressive Integrated Moving Average (ARIMA) models in order to strengthen evaluations. This presentation focuses on how ARIMA models improve upon simpler pre-post designs. When randomization or control groups are not possible or appropriate, it is possible to use ARIMA models to determine attribution with a high leve l of confidence. Challenges and strategies for high quality ARIMA models are discussed. The presentation includes an example of conducting an intervention analysis using ARIMA modeling in SPSS.