Can you change my preferences? Effect of social influence on intertemporal choice behavior

When in Doubt, Follow the Crowd? Responsiveness to Social Proof Nudges in the Absence of Clear Preferences

“…It was shown that social proof information lead to higher contributions for employees who were already in the plan with a low contribution, but had no effect for those who were not already enrolled. When the social proof information does not match the (slight) preference it might be regarded as irrelevant, (however, see Calluso et al, 2017). We expected that the inconsequential task of color categorization would be sufficient to invoke indifference, however, it could be argued that true indifference was demonstrated by the participants who failed to observe the social proof bar in the nudge conditions.…”

Concreteness and emotional valence of episodic future thinking (EFT) independently affect the dynamics of intertemporal decisions

“…During the baseline intertemporal choice task, participants were asked to make a series of hypothetical (i.e., subjects did not receive any payment for their participation in the study) choices between an immediate fixed amount of money (i.e., 10€) and a delayed one. The latter option was parametrically manipulated across seven amounts (i.e., 15€, 25€, 30€, 40€, 45€, 55€ and 60€) and six waiting times (i.e., 7, 15, 30, 60, 90, and 180 days), thus obtaining 42 different choice pairs (see [42,45,46] for a similar task design; Fig 1A). Each choice pair was repeated five times for each condition, thus resulting in a total of 210 trials.…”

“…Using the subjects’ observed behaviour and the well-known hyperbolic function as a model to describe the decline of subjective value with increasing time delays [54,55], we estimated the subject-specific discount rates ( k parameter), in line with a standard routine also employed in a series of previous studies (see [16,18,42,45,46,56]). As a first step, we calculated–for each time delay—the fraction of times in which participants selected the future reward over the immediate one as a function of the objective amount of the delayed option.…”

“…In the second analysis the same effect was tested on the distribution of choices for the two response types (i.e., now, later (see S1 File, S1 Fig). The third analysis, instead, which was conducted on the subjective values, was specifically aimed at testing whether the effect of the EFT manipulation was differential for choices associated with short- vs. long-time intervals (see also [46] for previous results from our group on this question). In this latter analysis, the original six time delays were collapsed into a short and a long time delay (i.e., short: 7, 15 and 30 days; long: 60, 90 and 180 days) and subjective values were predicted using an LMM with condition (baseline, negative, neutral and positive), time delay (short and long) and their interaction as fixed effects.…”

“…A third set of analyses was finally aimed at testing whether the magnitude of the effect associated with the EFT/emotional valance manipulation was differential across individuals with different baseline discounting preferences (i.e., discounters vs. farsighted). More specifically, in line with the methodology employed in our previous studies [42,46], this analysis was conducted on both explicit and implicit measures and was performed by dividing our sample in discounters and farsighted individuals by means of a median split of the baseline discount rates (i.e., discounters: k > median; farsighted k < median, p < 0.001). In the first analysis, an index of the magnitude of the EFT manipulation for each emotional valence (i.e., k shift ) was computed, which was derived from the difference between the baseline discount rate ( k Base ) and the discount rates associated with the three emotional valences obtained during the EFT session [i.e., K shift neg = K base −k neg ; K shift neu = K base −k neu ; K shift pos = K base −k pos ].…”