Acceso usuarios
Andrea Révész 
This commentary discusses some methodological innovations and challenges in testing two cognitive models of task-based instruction and learning: Robinson's Cognition Hypothesis (2001) and Skehan's Trade-Off Hypothesis (1998). My focus, in view of these models' key constructs, is the extent to which current methodological approaches are adequate to address the theoretical questions posed. I argue that there is a need for more extensive use of methods that can provide independent evidence of construct validity for the independent variables and causal processes invoked by the frameworks. By exploring some innovative task-based research and drawing on examples from other fields, I offer suggestions for improving our practices in future research.
