Research on learning from solving transformational problems has shown that the extent to which a goal is clearly specified to a problem solver as a problem state affects the problemsolving strategy used. Transformational problems are characterized by an initial problem state, a goal state, and a set of operators to transform the initial problem state into the goal state. Under goal-specific conditions novice problem solvers work backward from the goal setting subgoals until equations containing no unknowns other than a desired goal state are encountered (i.e., means-ends analysis). Under nonspecific goal conditions novice problem solvers work forward attaining the desired goal by choosing equations which allow a value for an unknown to be calculated (i.e., history-cued strategy). The goal-free effect refers to the finding that practicing by solving problems with a nonspecific goal imposes a lower cognitive load and leads to better learning than practicing by solving problems with a specific goal.