Is intertemporal decision-making based on a temporal discounting process? Evidence from double-dissociation paradigm
Yang-Yang ZHANG,Shu LI,Lei ZHOU,XuQun YOU,Zhu-Yuan LIANG
DOI: https://doi.org/10.1360/n052017-00104
2018-05-01
Scientia Sinica Vitae
Abstract:Intertemporal decision-making (IDM) refers to tradeoffs among options occurring at different time points. The mainstream theories of IDM can be divided into two categories according to the theories’ basic hypothesis on cognitive processes, namely, whether IDM is grounded on a discounting computation process or not. The two categories are discounting and non-discounting models. However, due to the lack of reliable and effective experimental paradigms, only a few studies have directly examined whether this hypothesis is true or not. To determine if IDM is based on the discounting computation process, this study designed a new double-dissociation experimental paradigm that includes a pair of tasks, namely, a baseline discounting task, in which participants make choices by following the discounting computation process, and the free-choice intertemporal task, in which participants make choices on their own. By adopting a double-dissociation logic, this study selected a pair of manipulating variables (computational difficulty and magnitude of outcomes) and moderator variables (arithmetic ability and cognitive reflection ability) that affect only the baseline discounting task or free-choice intertemporal task. Subsequently, task performances and the hierarchical Bayesian modeling results of the pair of tasks were compared at the level of main tasks and manipulating and moderator variables. Results showed that the performance of the baseline discounting task was in line with the hypothesis of discounting computation. As computational difficulty increased, participants required increased reaction time to complete the baseline discounting task, and their task accuracy was affected by computational difficulty and arithmetic ability. The performance for the free-choice intertemporal task differed from that for the baseline discounting task. Compared with the scenario for the baseline discounting task, participants required less time to complete the free-choice intertemporal task, and they selected less sooner and smaller (SS) options when the outcome magnitude was large. Moreover, the results of moderator variables and model fitting exhibited double dissociation effects. Arithmetic ability only influenced the participants’ choice in the baseline discounting task, and cognitive reflection ability only influenced the participants’ preference in the free-choice intertemporal task. Meanwhile, computational difficulty affected only the model fitting for the baseline discounting task, and magnitude of outcomes affected only the model fitting in the free-choice intertemporal task. The results indicated that the new experimental paradigm was effective in testing the discounting computation process hypothesis of intertemporal choice. Furthermore, the free-choice intertemporal task might be not in line with the prediction of the discounting computation hypothesis on task performances, moderator variables, and model fitting. This study highlights a new methodological exploration for research on the intertemporal choice process.