Research shows that computer-generated corrective feedback can promote second language development, but there is no consensus about which type is the most effective. The scale is tipped in favor of more explicit feedback that provides metalinguistic explanations, but counterevidence indicates that minimally explicit feedback of the "right/wrong" type may promote comparable learning outcomes. Addressing these conflicting findings, the present study investigated the effects of different types and amounts of practice as variables that may moderate the effectiveness of computerized "right/wrong" feedback. Fifty-two learners of intermediate Spanish completed either 28 or 56 items of an input-based task with 2 or 4 options targeting Spanish past counterfactual conditional sentences. Quantitative results on achievement scores showed that differences in amount of practice might contribute to explaining the conflicting findings in the literature. Additionally, a qualitative analysis of participant mouse-click histories illustrated participant use of elimination strategies to redefine the 4-option tasks, while participant think-alouds revealed the increased boredom and fatigue induced by the extra amount of practice. This study thus contributes to the debate on the effects of different types of computerized feedback and the development of hybrid and online language learning programs, while underscoring the importance of triangulating data from multiple sources.