Dynamic integration of forward planning and heuristic preferences during multiple goal pursuit

Selecting goals and successfully pursuing them in an uncertain and dynamic environment is an important aspect of human behaviour. In order to decide which goal to pursue at what point in time, one has to evaluate the consequences of ones actions over future time steps by forward planning. However, when the goal is still temporally distant, detailed forward planning can be prohibitively costly. One way to select actions at minimal computational costs is to use heuristics. It is an open question how humans mix heuristics with forward planning to balance computational costs with goal reaching performance. To test a hypothesis about dynamic mixing of heuristics with forward planning, we used a novel stochastic sequential two-goal task. Comparing participants decisions with an optimal full planning agent, we found that at the early stages of goal-reaching sequences, in which both goals are temporally distant and planning complexity is high, on average 42% (SD = 19%) of participants choices deviated from the agents optimal choices. Only towards the end of the sequence, participants behaviour converged to near optimal performance. Subsequent model-based analyses showed that participants used heuristic preferences when the goal was temporally distant and switched to forward planning when the goal was close. Author summaryWhen we pursue our goals, there is often a moment when we recognize that we did not make the progress that we hoped for. What should we do now? Persevere to achieve the original goal, or switch to another goal? Two features of real-world goal pursuit make these decisions particularly complex. First, goals can lie far into an unpredictable future and second, there are many potential goals to pursue. When potential goals are temporally distant, human decision makers cannot use an exhaustive planning strategy, rendering simpler rules of thumb more appropriate. An important question is how humans adjust the rule of thumb approach once they get closer to the goal. We addressed this question using a novel sequential two-goal task and analysed the choice data using a computational model which arbitrates between a rule of thumb and accurate planning. We found that participants decision making progressively improved as the goal came closer and that this improvement was most likely caused by participants starting to plan ahead.