Examining the Reliability of Running Records: Attaining Generalizable Results