A series of studies by Wulf and colleagues have found robust effects in support of an external, over an internal focus of attention when teaching complex motor skills (e.g., Wulf, et al., 2000; Wulf, Höß, & Prinz, 1998; Wulf, Lauterbach, & Toole, 1999; Wulf, Shea, & Whitacre, 1998). The present researchers and colleagues at The University of Virginia, have attempted to replicate and extend research by Wulf with four previous studies, none of which found support for difference between internal and external focus of attention. All of these studies and Wulf’s have used absolute error (AE) as the form of performance error measure. In the current study, a second dependent measure, variable error (VE) was utilized to measure learners’ consistency, rather than solely measuring absolute error (AE). Surprisingly, significant results were found in the retention and transfer tests using VE, but were not found using AE. These results, lead the current researches to contend that a consistency measure (VE) may in fact be a more powerful and effective form of measure when assessing the acquisition of complex tasks. Participants (N=28) volunteered for three 30-minute sessions. A shuffleboard task was chosen as a novel task, and participants were assigned either an internal, or external focus of attention. The task was to push the shuffleboard disc 20 feet, with the stick in the non-dominant arm. Both positive and negative error (AE) from the target lines were measured in inches, on the linear target. For both absolute error and variable error, a 2 (group: internal or external focus) x 3 (pre-test, post-test, retention test) ANOVA was performed, with a one-way ANOVA analyzed separately on the transfer test. At the alpha level of .05, no significant differences were found at any point (pre-test, post-test, retention or transfer) using AE measures; however, there were significant differences found in the VE measures on the retention test (p=.044), and transfer test (p=.035), favoring the external focus of attention. While AE is an important measure for instructors to help learners eliminate errors, it is equally important to measure participants’ performance consistency (VE). Absolute error (AE) will remain an essential measure, but the current study reinforces that multiple measures may give a better picture of the learning process.