RESUMEN
Errorless learning (EL) is an instructional procedure involving error reduction during learning. Errorless learning is mostly examined by counting correctly executed task steps or by rating them using a Task Performance Scale (TPS). Here, we explore the validity and reliability of a new assessment procedure, the core elements method (CEM), which rates essential building blocks of activities rather than individual steps. Task performance was assessed in 35 patients with Alzheimer's dementia recruited from the Relearning methods on Daily Living task performance of persons with Dementia (REDALI-DEM) study using TPS and CEM independently. Results showed excellent interrater reliabilities for both measure methods (CEM: intraclass coefficient [ICC] = .85; TPS: ICC = .97). Also, both methods showed a high agreement (CEM: mean of measurement difference [MD] = -3.44, standard deviation [SD] = 14.72; TPS: MD = -0.41, SD = 7.89) and correlated highly (>.75). Based on these results, TPS and CEM are both valid for assessing task performance. However, since TPS is more complicated and time consuming, CEM may be the preferred method for future research projects.