The variability effect occurs when learners' exposure to highly variable tasks results in better learning. It was hypothesised that learners who studied high variability worked examples would obtain higher post-test scores compared to learners who studied low variability examples, and learners who self-generated problem solutions for the same high or low variability tasks. This hypothesis was not supported. However, subjective ratings of difficulty supported the assumptions based on cognitive load theory. From a practical perspective, these results suggest that novice learners should initially be presented with low variability, high-guidance learning tasks to reduce a potential cognitive overload.