The underlying mechanisms of choice-induced preference change effects

© 2018 Dr Katharina VoigtProminent models of decision making claim that choices are based on stable preferences. This thesis investigated whether preferences can also be shaped endogenously by the process of making hard choices. The experiments of this thesis were designed to address two main research questions. First, can decisions among equally preferred choice alternatives shape preferences. Second, if so, what are the underlying mechanisms of such choice-induced preference change effects. The first study thought to address whether choice-induced preference changes exist by adopting a novel incentivised free choice task, which eliminated methodological issues inherent in previous research. In this task, participants valued snack foods before making binary choices between equally valued items, and subsequently valuating the foods again. Additionally, participants completed an experimental control sequence, which assessed changes in valuations attributable to regression to the mean. Across a series of experiments, this task comprised of choices and valuations that were either realistic, incentive-compatible or hypothetical, incentive-incompatible. Preference changes increased for chosen and decreased for rejected alternatives, but only in decision scenarios that were realistic and incentive-compatible. This pattern of results suggests that preferences can change endogenously, that is, in the absence of external feedback or information, but rather as a function of previous relevant choices. The second study investigated the underlying neural and cognitive mechanisms of the effects demonstrated in Study 1 by means of functional magnetic resonance imaging (fMRI) and eye-tracking. Previous explanations of the choice-induced preference change effect claim that preferences change after difficult choices were made to adjust one’s preferences to previous decisions, and only if those could be remembered during revaluation. Our results suggest a radically different explanation. We show that preferences are already adjusted during the process of making hard choices: Fixation duration were predictive of upcoming choices as well as of the subsequent preference adjustments. These in-decisional changes in preferences were also reflected in activity in the dorsolateral prefrontal cortex and precuneus. These preference adjustments were measured at the time of re-evaluation, but only for choice outcomes that were remembered and were associated with hippocampus activity. These findings strongly refute previous post-decisional explanations of choice-induced preference change effects, but instead suggest that preferences evolve dynamically as decisions arise, potentially as a mechanism to prevent stalemate situations in underdetermined decision scenarios. The final study was designed to explain individual differences in how choice-induced preference change effects are expressed by means of resting state fMRI. The results revealed that individual differences in choice-induced preference change effects were explained by local and global fluctuations in low-frequency blood oxygen level-depended signalling within brain areas that were found to be associated with the effect in the second study. Specifically, neural signals in and coupling between brain areas associated with episodic memory (i.e. precuneus and hippocampus) and value-related processing (i.e. ventromedial prefrontal cortex) explained a significant fraction of the behavioural variability of choice-induced preference change effects. These results shift the focus further away from post decisional explanations. Rather they suggest that the functional configuration of these brain areas determines how individuals’ preferences shift in hard decision scenarios. In summary, the experiments presented in this thesis provide evidence that preferences are highly dynamic and can be shaped by the process of making hard choices. This formation process might depend on attentional shifts via eye-movements during the decision-making process as well as the initial functional configuration of brain areas associated with the effect. Taken together, these findings support the notion that our preferences evolve endogenously during underdetermined decision environments, which may act as an adaptive problem-solving mechanism