Bentley et al. have created a simple map-like model of selected determinants of human decision-making behavior, focusing attention on two factors: (1) the degree to which a decision is independent versus socially influenced (axis 1 of the map), and (2) the degree of transparency concerning the consequences of various decisions in terms of payoff and/or risk (axis 2). We agree that these factors play an important role in human decision-making processes, but we are not convinced that the map model of Bentley et al. will allow us, as stated by its authors, to “capture the essence of decision making” (target article, Abstract). Below we discuss the main shortcomings of their model.
Many aspects of the map model of Bentley et al. are simplified, some of them deliberately. The authors argue that “the map requires a few simplifying assumptions to prevent it from morphing into something so large that it loses its usefulness for generating potentially fruitful research hypotheses” (target article, sect. 2, para. 7). However, such an approach raises a question as to what degree such a simplified model still provides a useful representation of the analyzed phenomena. The authors seem to share these doubts, as they recommend further research to explore the effects of factors deliberately disregarded in their model.
Extensive research in the field of behavioral biology and social psychology has yielded numerous data demonstrating that social influences mediating choice behavior of animals and humans show a high degree of qualitative diversity. Bentley et al. are aware of the existence of various subcategories of social influences and name some of them (copying, verbal instruction, imitation). However, they quantify the impact of social influences on decision making in a very simplified way, solely by providing information on the degree to which a decision is socially influenced. Qualitative differences between various subcategories of social influences are disregarded. As a consequence, a decision half-influenced by imitation of decisions of other agents (“herdlike” behavior) and a decision influenced to the same degree by avoidance of imitation (nonconformist behavior) would occupy the same position on their map. Distinctions between different strategies employed in decision making are thus blurred instead of being emphasized. We may also note that the authors tend to identify social linking with herdlike behavior and pay little or no attention to social interactions leading to enhanced diversity of behavior.
Yet another set of problems is related to the fact that the map model proposed in the target article represents a continuous space defined by the two analytical dimensions, but the authors have divided it into four quadrants “for ease of discussion and application to example datasets” (target article, sect. 2, para. 7), and throughout their article they discuss various social phenomena mostly by assigning them to one of these quadrants, without any attempt to identify their exact position on the map. So, ultimately, they mostly use a discrete 2×2 matrix and not a continuous map. Such simplification has, however, some merits, too, as it puts aside the need to quantify precisely the analyzed phenomena by assigning to them the values of variables representing the two axes of the map, and it is far from obvious how to make the measurements of values of these variables or even to provide reliable estimations of them. However, that question must find a satisfactory solution if Bentley et al.’s map model is to be assigned to the domain of empirical science and give rise to testable hypotheses.
The authors also pay no attention to the fact that the degree of transparency concerning the consequences of various decisions is socially influenced. Therefore, the two axes of the map model of Bentley et al. are not independent: social influences may affect the choice process in two ways, directly (axis 1) and indirectly (axis 2). This raises an additional difficulty in assigning values to variables forming the axes of the map model. We should also bear in mind that information obtained from other individuals may enhance the transparency concerning the consequences of decisions, but it may also be involuntarily or purposefully misleading (the phenomenon of cheating).
Finally, the map model of causation of decision-making processes takes into account only causal factors related to the individual and/or social level of organization. Phenomena and processes taking place on lower levels of organization are disregarded. However, factors selected by Bentley et al. to explain the causation of human decision making processes exert their influence via modifications of neurobiological processes taking place in the agent's brain. Our knowledge about the neurobiological basis of animal and human choice behavior is already quite advanced. We are convinced that the processes of decision making could be much more profoundly understood if such issues as the contextual modulation of behavioral choice (including, in particular, the relative contribution of conscious processes vs. processes induced by subliminal stimuli) (Block Reference Block2007; Palmer & Kristan Reference Palmer and Kristan2011), the role of particular brain structures and neurotransmitter/neuromodulatory systems (Forbes & Grafman Reference Forbes and Grafman2013; Jung et al. Reference Jung, Sul and Kiu2013; Pleger & Villringer Reference Pleger and Villringer2013; Yu & Dayan Reference Yu and Dayan2005), and the role of synchronization of electrical activity rhythms in various parts of the brain (Guitart-Masip et al. Reference Guitart-Masip, Barnes, Horner, Bauer, Dolan and Duzel2013) were taken into account. In our opinion, Bentley et al.’s approach would be more successful if it were supplemented by information showing how the impact of factors representing analytical dimensions of their map is translated into neurobiological processes underlying decision making. Further refinements providing additional information on neurobiological correlates of human choice processes might make their map model still more useful and realistic.
You have
Access
Bentley et al. have created a simple map-like model of selected determinants of human decision-making behavior, focusing attention on two factors: (1) the degree to which a decision is independent versus socially influenced (axis 1 of the map), and (2) the degree of transparency concerning the consequences of various decisions in terms of payoff and/or risk (axis 2). We agree that these factors play an important role in human decision-making processes, but we are not convinced that the map model of Bentley et al. will allow us, as stated by its authors, to “capture the essence of decision making” (target article, Abstract). Below we discuss the main shortcomings of their model.
Many aspects of the map model of Bentley et al. are simplified, some of them deliberately. The authors argue that “the map requires a few simplifying assumptions to prevent it from morphing into something so large that it loses its usefulness for generating potentially fruitful research hypotheses” (target article, sect. 2, para. 7). However, such an approach raises a question as to what degree such a simplified model still provides a useful representation of the analyzed phenomena. The authors seem to share these doubts, as they recommend further research to explore the effects of factors deliberately disregarded in their model.
Extensive research in the field of behavioral biology and social psychology has yielded numerous data demonstrating that social influences mediating choice behavior of animals and humans show a high degree of qualitative diversity. Bentley et al. are aware of the existence of various subcategories of social influences and name some of them (copying, verbal instruction, imitation). However, they quantify the impact of social influences on decision making in a very simplified way, solely by providing information on the degree to which a decision is socially influenced. Qualitative differences between various subcategories of social influences are disregarded. As a consequence, a decision half-influenced by imitation of decisions of other agents (“herdlike” behavior) and a decision influenced to the same degree by avoidance of imitation (nonconformist behavior) would occupy the same position on their map. Distinctions between different strategies employed in decision making are thus blurred instead of being emphasized. We may also note that the authors tend to identify social linking with herdlike behavior and pay little or no attention to social interactions leading to enhanced diversity of behavior.
Yet another set of problems is related to the fact that the map model proposed in the target article represents a continuous space defined by the two analytical dimensions, but the authors have divided it into four quadrants “for ease of discussion and application to example datasets” (target article, sect. 2, para. 7), and throughout their article they discuss various social phenomena mostly by assigning them to one of these quadrants, without any attempt to identify their exact position on the map. So, ultimately, they mostly use a discrete 2×2 matrix and not a continuous map. Such simplification has, however, some merits, too, as it puts aside the need to quantify precisely the analyzed phenomena by assigning to them the values of variables representing the two axes of the map, and it is far from obvious how to make the measurements of values of these variables or even to provide reliable estimations of them. However, that question must find a satisfactory solution if Bentley et al.’s map model is to be assigned to the domain of empirical science and give rise to testable hypotheses.
The authors also pay no attention to the fact that the degree of transparency concerning the consequences of various decisions is socially influenced. Therefore, the two axes of the map model of Bentley et al. are not independent: social influences may affect the choice process in two ways, directly (axis 1) and indirectly (axis 2). This raises an additional difficulty in assigning values to variables forming the axes of the map model. We should also bear in mind that information obtained from other individuals may enhance the transparency concerning the consequences of decisions, but it may also be involuntarily or purposefully misleading (the phenomenon of cheating).
Finally, the map model of causation of decision-making processes takes into account only causal factors related to the individual and/or social level of organization. Phenomena and processes taking place on lower levels of organization are disregarded. However, factors selected by Bentley et al. to explain the causation of human decision making processes exert their influence via modifications of neurobiological processes taking place in the agent's brain. Our knowledge about the neurobiological basis of animal and human choice behavior is already quite advanced. We are convinced that the processes of decision making could be much more profoundly understood if such issues as the contextual modulation of behavioral choice (including, in particular, the relative contribution of conscious processes vs. processes induced by subliminal stimuli) (Block Reference Block2007; Palmer & Kristan Reference Palmer and Kristan2011), the role of particular brain structures and neurotransmitter/neuromodulatory systems (Forbes & Grafman Reference Forbes and Grafman2013; Jung et al. Reference Jung, Sul and Kiu2013; Pleger & Villringer Reference Pleger and Villringer2013; Yu & Dayan Reference Yu and Dayan2005), and the role of synchronization of electrical activity rhythms in various parts of the brain (Guitart-Masip et al. Reference Guitart-Masip, Barnes, Horner, Bauer, Dolan and Duzel2013) were taken into account. In our opinion, Bentley et al.’s approach would be more successful if it were supplemented by information showing how the impact of factors representing analytical dimensions of their map is translated into neurobiological processes underlying decision making. Further refinements providing additional information on neurobiological correlates of human choice processes might make their map model still more useful and realistic.