Dissociating the impact of movement time and energy costs on decision-making and action initiation in humans

Abstract

Recent theories and data suggest that adapted behavior involves economic computations during which multiple trade-offs between reward value, accuracy requirement, energy expenditure and elapsing time are solved so as to obtain rewards as soon as possible while spending the least possible amount of energy. However, the relative impact of movement energy and duration costs on perceptual decision-making and movement initiation is poorly understood. Here, we tested 31 healthy subjects on a perceptual decision-making task in which they executed reaching movements to report probabilistic choices. In three distinct blocks of trials, the reaching time and energy costs were independently varied while decision difficulty was maintained similar at the block level. Participants also performed a fully instructed delayed-reaching (DR) task in each motor condition. Results in that DR task show that time-consuming movements extended reaction times (RTs) in most subjects, whereas energy-consuming movements led to mixed effects on RTs. In the choice task, about half of the subjects decreased their decision durations (DDs) in the time consuming condition, while the impact of energy costs on DDs were again mixed across subjects. Decision accuracy was overall similar across motor conditions. These results indicate that movement duration and, to a lesser extent, energy expenditure, idiosyncratically affect perceptual decision-making and action initiation. We propose that subjects who shortened their decisions in the time consuming condition of the choice task did so to limit a drop of their rate of reward.

Movement characteristics impact decision-making and vice versa

“…At first glance, our findings seem to be at odds with results of Reynaud and colleagues (2020), who also investigated how movement duration affects decision-making. In contrast to our findings, faster as compared to slower movements were associated with longer decisions (see also Lunazzi et al, 2021). However, in their task 160 correct decisions completed an experimental block, and faster movements allowed to finish trials faster.…”

“…Therefore, success probability of each decision can be formally calculated as the proportion of all hypothetical growth patterns of bananas following the decision in which the chosen banana wins. In accordance with previous work (Derosiere et al, 2019; Reynaud et al, 2020; Saleri Lunazzi et al, 2021; Thura & Cisek, 2014, 2016, 2017), success probability can therefore be computed as where the probability p of the selected banana S winning is equivalent to the probability of unselected banana N not winning. N N is the number of frames in favor of the unselected banana, N O is the number of outstanding frames, both at the estimated time of decision offset (i.e.…”

“…Sixty-two participants performed two separate sessions of a behavioral online experiment from home. We administered an adaptation of the Tokens task, which has been widely used to investigate the effect of time pressure on sensorimotor processes in humans (Cisek et al, 2009; Derosiere et al, 2019; Reynaud et al, 2020; Lunazzi et al, 2021; Thura, 2020). Each trial consisted of a dedicated decision and movement phase in order to measure the time participants would allocate to deciding and moving (see Figure 1).…”

“…In line with this, drift rate was significantly reduced in fast-decision blocks as compared to slow-decision blocks, suggesting that deciding with limited evidence and under time pressure may have been more difficult, and hence more effortful, rather than less effortful. In addition, a recent study explicitly tested in a related task whether movement effort is compensated by shortening decision duration, finding no such effect with 31 participants (Lunazzi et al, 2021). Given these considerations, we interpret current findings as reflecting inherent properties of the sensorimotor system rather than strategic adaptations with changing task contexts.…”

“…Endpoint measures were computed which reflect decision and motor performance: Decision duration was computed for each participant as average reaction times in each experimental condition minus the average reaction time in the corresponding SRT block (c.f. Derosiere et al, 2019; Reynaud et al, 2020; Lunazzi et al, 2021; Thura, 2020; Thura & Cisek, 2014, 2016, 2017). This ensured that only the time taken to decide was compared between experimental blocks, without confounding it with the time needed to perform a first finger tap.…”