Reviewing the impact of problem structure on planning: A software tool for analyzing tower tasks

Abstract

Cognitive, clinical, and neuroimaging studies on planning abilities most frequently implement the Tower of London task or one of its variants. Yet, cumulating evidence from a series of experiments suggests that the commonly used approximation of problem difficulty in terms of the minimum number of moves for goal attainment is too coarse a measure for the underlying cognitive operations, and in some cases may be even misleading. Rather, problem difficulty can be more specifically characterized by a set of structural task parameters such as the number and nature of optimal and suboptimal solution paths, the required search depths, the patterns of intermediate and goal moves, goal hierarchies and the associated degree of ambiguity in the sequential ordering of goal moves. First applications in developmental and patient studies have proven fruitful in targeting fundamental alterations of planning abilities in healthy and clinical conditions. In addition, recent evidence from neuroimaging shows that manipulations of problem structure relate to separate cognitive and neural processes and are accompanied by dissociable brain activation patterns. Here, we briefly review these structural problem parameters and the concepts behind. As controlling for task parameters and selecting a balanced problem set is a complex and error-prone endeavor, we further present TowerTool, a software solution that allows easy access to in-depth analysis of the problem structure of widely used planning tasks like the Tower of London, the Tower of Hanoi, and their variants. Thereby, we hope to encourage and facilitate the implementation of structurally balanced task sets in future studies on planning and to promote transfer between the cognitive, developmental, and clinical neurosciences.