The complementary role of affect-based and cognitive heuristics to make decisions under conditions of ambivalence and complexity

Abstract

Little is known about the interplay between affective and cognitive processes of decision making within the bounded rationality perspective, in particular for the debate on adaptive decision making and strategy selection. This gap in the knowledge is particularly important as affect and deliberation may direct preferences in opposite directions. How do decision makers solve such dissonance? In this paper, we address this question by exploring the use of integral affect as a choice heuristic in comparison with and in conjunction to “take the best,” and weighted addition of attributes (WADD). We operationalize theories of reliance on affect in choice through a "Take the emotionally best" algorithm. Its predictive power is experimentally tested against other models, including mixed-sequential cognitive/affective procedures. We find that individual decisions are better predicted by a sequential combination of "Take the emotionally best" and "Take the best" with a slight dominance of the former. Conditions of cognitive/affective ambivalence, low discrimination ability and high complexity provide the cognitive architecture where such blended choice strategies predict decisions more precisely. This implies that reliance on integral affect may precede the use of cognitive cues following an ecological rationality perspective rather than supporting a kind of competition between affect and cognition as implied in current literature.

Incorporating emotion into cue-based political judgment modeling.

“…Adding to prior research (Trujillo, 2018) which suggests that experienced emotion predicts judgment and decision-making more accurately than compensatory and noncompensatory cognitive models, the valence of predicted affective forecasts will better predict policy decision-making than will cue-based modeling regardless of the type of model (i.e., fast and frugal heuristics or full compensatory models).…”

“…We calculated separate cue utilization scores for each participant and display them along with their measurement properties in Table 3. Experiments which compare compensatory and noncompensatory models routinely use participant ratings to ascertain cue values, cue weights, and cue utilization scores (e.g., Dhami & Harries, 2001; Garcia-Retamero & Dhami, 2009; Trujillo, 2018; Wang, 2008).…”

Is Reliance on the Affect Heuristic Associated With Age?

“…The lack of cross-task convergence in affect heuristic use could have been influenced by methodological factors as well: Although all tasks were completed within a single test session by the same participants, indices derived from different affect heuristic tasks did not converge. Conceivably, reliance on the affect heuristic may be context-or stimulus-dependent rather than a stable inter-individual trait: As is the case for other heuristics, use of the affect heuristic increases under time pressure, when cognitive resources run low, or in the absence of explicit information (Fleming, Townsend, van Hilten, Spence, & Ferguson, 2012;Finucane et al, 2000;King & Slovic, 2014;Shiv & Fedorikhin, 1999;Trendel & Werle, 2016;Trujillo, 2018). For example, decision makers are more likely to choose emotionally favorable options over less favorable options when their cognitive load is high and the choice environment is complex (Shiv & Fedorikin, 1999;Trujillo, 2018).…”

“…Conceivably, reliance on the affect heuristic may be context-or stimulus-dependent rather than a stable inter-individual trait: As is the case for other heuristics, use of the affect heuristic increases under time pressure, when cognitive resources run low, or in the absence of explicit information (Fleming, Townsend, van Hilten, Spence, & Ferguson, 2012;Finucane et al, 2000;King & Slovic, 2014;Shiv & Fedorikhin, 1999;Trendel & Werle, 2016;Trujillo, 2018). For example, decision makers are more likely to choose emotionally favorable options over less favorable options when their cognitive load is high and the choice environment is complex (Shiv & Fedorikin, 1999;Trujillo, 2018). As a result, use of the affect heuristic may be sensitive to the specifics of the task design.…”