Eyes on the Prize? Evidence of Diminishing Attention to Experienced and Foregone Outcomes in Repeated Experiential Choice

Abstract

Recently there has been increased interest in decisions-from-experience (where decision makers learn from observing the outcomes of previous choices), which provide valuable insights into the learning and preference construction processes underlying many daily decisions. Several process models have been developed to capture these processes, and while such models often fit the data well, many assume that the decision maker is a vigilant observer, processing each outcome. In two studies, we provide a critical test of this assumption using eye tracking to record directed visual attention when participants choose repeatedly among two options, each time being shown the outcome for their chosen option and for the foregone option. Consistently, we find that the vigilance assumption is not supported, with decision makers often not attending to outcome information. Moreover, (in)attention to outcomes is predictable, with vigilance decreasing as more choices are made, and being greater for obtained than for foregone outcomes, and when options deliver only gains as opposed to losses, or a mixture of gains and losses. Furthermore, we find that this variation in attentional allocation plays a central role in the apparent indecisiveness (inconsistency) in choice, with increased attention to foregone outcomes predicting switches to that option on the next choice. Together, these findings highlight the value of eye tracking in investigations of decisions-from-experience, providing novel insight into cognitive processes underlying them.

Choice in experiential learning: True preferences or experimental artifacts?

“…Participants rarely converge to selecting the same option exclusively. Specifically, while participants in DF tasks often learn to select options of higher expected value (EV) more frequently with experience (Ashby & Rakow, 2015;Jessup, Bishara, & Busemeyer, 2008), some variability in choice remains even after a great deal of experience (Konstantinidis, Ashby, & Gonzalez, 2015), and even occurs when options are fully described (Barron & Lieder, 2010; Lejaraga & Gonzallez, 2011; Weiss-Cohen, Konstantinidis, Speekenbrink, & Harvey, 2016;Yechiam, Barron, & Erev, 2005). As noted above, this pattern was observed in probability learning tasks (Bereby-Meyer & Erev, 1998;Erev, Bereby-Meyer, & Roth, 1999) and in a DF task where one option deterministically dominated another option: Variance in choice remained after 200 trials of choosing between one option providing 11 points with certainty and another providing 10 points with certainty (Haruvy & Erev, 2002).…”

The Influence of Time Estimation and Time‐Saving Preferences on Learning to Make Temporally Dependent Decisions from Experience

“…The current studies relate to theory regarding experiential learning . As is often the case in decisions-from-experience (Ashby & Rakow, 2016) and studies of temporal perception (Wright et al, 1997), participants were able to improve their performance with experience. This similarity with previous findings is interesting given that the current studies departed from the paradigm of recent decisions-from-experience tasks where rewards are independent of when an action is taken.…”

“…More importantly, the current research suggests that the ability to learn was greatly constricted by the confines of temporal perception, which indicates the importance of understanding the limitations of cognitive systems when examining human behavior in experiential learning (Ashby & Rakow, 2016;Rakow, Newell, & Zougkou, 2010). Specifically, when decisions were reliant on temporal estimation, there was an incentive for minimizing wait.…”

“…In simple binarychoice paradigms of decisions-from-experience, where explicit information about potential outcomes and probabilities is hidden, individuals learn the values of options by repeatedly selecting from them and observing their outcomes (Barron & Erev, 2003;Hertwig & Erev, 2009). In decisionsfrom-experience, individuals tend to maximize (make more choices for the option of highest expected value) more with experience (Ashby & Rakow, 2016;Jessup, Bishara, & Busemeyer, 2008). However, when the probability of receiving a reward is highly variable or small, increased maximization does not always occur (Barron & Erev, 2003;Mehlhorn, Ben-Asher, Dutt, & Gonzalez, 2014).…”

“…However, when the probability of receiving a reward is highly variable or small, increased maximization does not always occur (Barron & Erev, 2003;Mehlhorn, Ben-Asher, Dutt, & Gonzalez, 2014). Interestingly, Ashby and Rakow (2016) reported that in a decisions-fromfeedback paradigm (where repeated consequential choices with feedback are made), attention to outcomes diminished with experience. This finding indicates that decision makers sought to minimize their cognitive investment in the task.…”