Salience effects in information acquisition: No evidence for a top-down coherence influence

Abstract

The Integrated Coherence-Based Decision and Search (iCodes) model proposed by Jekel et al. (Psychological Review, 125 (5), 744-768, 2018) formalizes both decision making and pre-decisional information search as coherence-maximization processes in an interactive network. Next to bottom-up attribute influences, the coherence of option information exerts a top-down influence on the search processes in this model, predicting the tendency to continue information search with the currently most attractive option. This hallmark "attraction search effect" (ASE) has been demonstrated in several studies. In three experiments with 250 participants altogether, a more subtle prediction of an extended version of iCodes including exogenous influence factors was tested: The salience of information is assumed to have both a direct (bottom-up) and an indirect (top-down) effect on search, the latter driven by the match between information valence and option attractiveness. The results of the experiments largely agree in (1) showing a strong ASE, (2) demonstrating a bottom-up salience effect on search, but (3) suggesting the absence of the hypothesized indirect top-down salience effect. Hence, only two of three model predictions were confirmed. Implications for various implementations of exogenous factors in the iCodes model are discussed.

Keywords: Coherence; Decision making; Information search.

Fig. 1

Modeling exogenous influences on information search within iCodes. Links in the network can be excitatory (solid lines) or inhibitory (dashed lines) and bi-directional (two arrow heads) or uni-directional (one arrow head). a Original network without exogenous influences. b Exogenous influences modeled at the cue-value level. Here, salient cue values receive more activation directly. c Post hoc version of iCodes with exogenous influences modeled at the option level. In this version, the options that contain any salient information receive more activation

Fig. 2

Mean size of the attraction search score (ASS) as observed in the experiments for each experimental condition (a), predicted size of the ASS by using the a priori extended iCodes model (b), and predicted size of the ASS by the post hoc model (c). Error bars are standard deviations across the eight stimulus patterns

Fig. 3

Screenshots of trials in Experiment 2. In Phase 1, participants judged the relative attractiveness of both options by using the ruler. After the judgment, they could acquire additional cue information in Phase 2 (i.e., request information hidden behind one of the “?”)

Fig. 4

Violin plots showing the frequency distributions of the overall attraction search scores in the three experiments. Dashed line denotes independence of search and option attractiveness, higher values denote a stronger attraction search effect. Large dots denote median

Fig. 5

Attraction search scores as a function of predicted effect direction, and type of salience effect (direct vs. indirect) predicted by iCodes in the three experiments. Error bars denote 95% confidence interval based on within-subjects error

Fig. 6

Attraction search scores as a function of value of salient cue and option attractiveness. Error bars denote 95% confidence interval based on within-subjects error

Fig. 7

Predicted confidence ratings (i.e., differences in activations of option nodes) derived from the a priori extended iCodes model (x-axis) and mean observed confidence ratings (y-axis) for the eight stimulus patterns 1–8 (black dots) and 8 stimulus patterns × 7 salience conditions = 56 mean confidence ratings (grey dots). The regression line with indication of the 95% confidence interval is based on the confidence ratings of the eight stimulus patterns (i.e., black dots)

Fig. 8

Predicted confidence ratings (i.e., differences in activations of option nodes) derived from the a priori extended iCodes model (x-axes) and mean observed confidence ratings (y-axes) for the eight stimulus patterns plotted in displays 1–8 and seven salience conditions (dots within pattern displays) plotted in each display. The regression lines with indication of the 95% confidence intervals are based on the confidence ratings of the seven salience conditions for each pattern.