Judges’ Queries and Presenter’s Replies

  • May 20, 2013 | 05:10 p.m.

    Your video suggested that processes connected to the choice of effector or difficulty of delivering the response could affect the choice. Do you have evidence for this, and if not, how might you investigate this question?

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 20, 2013 | 08:22 p.m.

    Thank you for the apt question!

    As suggested in the video’s extreme example where an apple becomes less likely to be chosen if it requires much more physical effort to obtain than an orange, it appeals to our intuition that the difficulty of obtaining objects of value can affect choices about those objects. Whereas it is less apparent that the effectors involved in implementing a value-based decision can influence that decision, this must be the case when considering the cognitive and metabolic costs associated with using any effector, such as the left hand of a right-handed person.

    The notion that cognition and decision making are embodied (i.e., determined by the body and interactions with the physical world), which motivates this research, is not a particularly novel one (Cisek & Kalasaka, ARN, 2010). A considerable and growing body of evidence from neurophysiological data is consistent with this idea and suggests that the neural mechanisms for implementing motor responses are also critical for making decisions about which responses to execute, such as in the lateral intraparietal area implicated in saccades (Gold & Shadlen, ARN, 2007).

    The findings presented here provide further direct evidence for effectors influencing value-based decision making insofar as there are differences in both behavior and neural activity related to choices across the three experimental conditions—that is, when choosing with the hands, with the feet, or with the effectors unknown at the time of choice. In contrast, a serial model of decision making and subsequent action planning would instead predict that neural comparison processes are uniformly engaged regardless of the effectors necessary for a decision.

    I hope that addresses your question. Let me know if not.

  • May 21, 2013 | 08:36 a.m.

    Very interesting.
    I am wondering how your model/paradigm might extend to more complex choices (e.g., choices involving many tradeoffs), and similarly to choices/decision we rarely make. Would you expect the same pattern of results?

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 21, 2013 | 04:26 p.m.

    Great question, thank you!

    It is certainly feasible that a model in which valuation and comparison operate at these levels of representation also applies for more complex and less familiar choices as long as the necessary actions can be well defined. Choices between familiar foods are also perhaps more complex than they might appear at first. These decisions require integration of information about the organism’s internal state and value signals computed for various attributes of the foods and actions for obtaining them. As a result, people even make choices that are inconsistent with previous ones, such as choosing the apple one day and the orange the next.

    Nonetheless, people are also capable of making complex decisions at more abstract levels of representation, such as deciding on a job or a graduate program. Representations in the domains of space, effectors, and actions likely have less influence on valuation and comparison for these more abstract decisions, which are perhaps to some extent uniquely human. However, throughout the course of our evolution, the nervous system has primarily served to guide the organism’s interactions with the physical world and makes decisions with this core purpose. Thus, for example, even if the words one has to say in order to accept a job have little weight on her decision about the job, she is likely still (consciously or not) considering what she might say in either case while making the decision.

    I hope that addresses your question. Let me know if not.

  • May 21, 2013 | 03:38 p.m.

    You mentioned in your video that choices are rarely made by the foot, however when driving a vehicle, choices are routinely made by the foot (accelerate versus brake). There are accommodations (for some people who are unable to control a vehicle with their feet) that allow people to accelerate and brake using their hands, how might the change from foot to hand affect their brain and ultimately their choice when driving?

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 21, 2013 | 08:23 p.m.

    Thank you for your very thought-provoking question!

    To clarify, it is actually not the case that choices are rarely made by the foot. Rather, as mentioned, people just do not make these particular kinds of choices with the feet. Aside from driving, another familiar example of decision making involving these effectors is navigating while walking so as to avoid obstacles and reach some goal. One not only steps, but also decides when and where it is valuable to step.

    If the driver’s hands were used for both steering and controlling acceleration, I speculate that interference between decision-making processes related to each subtask would increase the difficulty of driving accurately because actions for implementing them would be considered simultaneously and impose cognitive demands on controlling the shared effectors. It seems feasible then that cars were designed to have separate roles for the hands and feet at least in part to avoid this interference. There would be less interference, however, if one were able to safely steer with one hand and control acceleration with the other because the neural representations of each hand are predominantly distinct and even lateralized.

    Moreover, if one were accustomed to controlling acceleration with the feet and began to drive using only the hands, initially it would be more difficult to make decisions about acceleration because of the unfamiliarity of the effectors, which implies a noisier representation providing less accurate information for decision-making processes.

    I hope that addresses your question. Let me know if not.

  • Icon for: Mary Gauvain

    Mary Gauvain

    Judge
    May 21, 2013 | 06:21 p.m.

    Your overarching framework connects motor activity and mental processing, can you explain the significance of this framework? In other words, why is it important to view mental processing in human beings as constrained or affected by physical actions?

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 21, 2013 | 11:43 p.m.

    Thank you for this important question!

    Recognizing the role of embodiment is crucial for a complete mechanistic understanding of how people think and make decisions to the extent that embodiment is a fundamental property of cognition and decision making. Almost every decision can be framed in terms of action selection because at its core decision making is about guiding the interactions of organisms with the physical world. An ecologically relevant perspective on decision making must then incorporate embodiment in order to account for how mobile people make decisions about how to act in dynamic environments, as opposed to treating the brain as an isolated, static unit limited to abstract thought.

    Attempting to understand how decisions are made without any regard for how they are implemented via motor output is unproductive because the implementation nearly always shapes the decision-making process at different levels of representation. Nonetheless, much of the research in this field paints an incomplete portrait that essentially neglects this significant aspect of how the nervous system actually makes decisions. The false assumption that I challenge is that actions are ancillary to decisions insofar as they are merely planned at the very end of a sequence of mental processes confined within the brain. Rather, decision making and motor activity are inextricably intertwined and can involve both feedforward and feedback signaling in parallel. Thus, neither phenomenon can be properly understood without taking the other into consideration.

    I hope that addresses your question. Let me know if not.

  • May 21, 2013 | 09:38 p.m.

    This is interesting. How might your research be applied? What does having a better model of embodied decision making do for society? Who would use this information, to what end?

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 22, 2013 | 02:10 p.m.

    Thank you for this much-needed question!

    Generally speaking, this research constitues a step towards achieving a complete mechanistic understanding of how people make and implement decisions. A better understanding of decision making invariably equips all of us to make better decisions to the extent that we are otherwise largely unaware of what we are actually doing when making decisions. More specifically, by emphasizing the significance of embodiment and levels of representation in decision making, this research contributes to an ecologically relevant perspective on decision making that can be directly applied to real-world scenarios. This perspective recognizes that the dynamics of a person’s movements and environment are critical variables in decision making. Considering that nearly all real-world decisions are implemented with actions in a dynamic world and that information about potential actions feeds into the decision-making process, no decision can be properly understood without regard for how the potential actions shape that decision.

    For example, one application of this research with immediate benefits for society is in improving treatment for substance dependence and compulsive overeating. These conditions are characterized by overpowering urges not only to consume a high-valued substance (e.g., a drug or food), but also to execute actions endowed with value by virtue of their association with obtaining and consuming the substance. Thus, therapies can be optimized for targeting high-valued but harmful actions leading up to consumption in addition to consumption itself. For example, a patient could be instructed to repeatedly prepare the addictive substance as she normally would and then expose herself to an aversive stimulus rather than an appetitive one (i.e., the substance) at the end of the sequence.

    I hope that addresses your question. Let me know if not.

  • May 22, 2013 | 09:27 p.m.

    Thanks Jaron, that does help.

  • Further posting is closed as the competition has ended.

Presentation Discussion

  • Icon for: Laurel James

    Laurel James

    Trainee
    May 21, 2013 | 12:55 a.m.

    Good job!

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 21, 2013 | 04:27 p.m.

    Thank you!

  • May 21, 2013 | 09:31 p.m.

    Cool experiment, Jaron. Makes me think I should train my feet to be better at making decisions! So what’s the goal of finding out this kind of information? More about understanding signals in the brain? Or psychology of decision-making? Something else?

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 22, 2013 | 12:50 p.m.

    Thank you!

    I would like to think that this is a decent example of how neuroscience informs psychology and vice versa. As a cognitive neuroscientist, I staunchly believe that one must understand the brain and nervous system in order to understand the mind and vice versa.

  • May 22, 2013 | 02:54 p.m.

    Nice video. How do you think this difference in decision-making generalizes beyond your study system?

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 22, 2013 | 04:52 p.m.

    Thank you!
    I think it is fair to suggest that levels of representation related to actions are implicated in nearly all decision making to the extent that nearly all decisions translate to actions. So when your ghost crab decides that it wants shelter, its nervous system is directly assigning value not only to the goal of shelter, but also to the subgoals involved in controlling its appendages to orchestrate the necessary burrowing movements. Great work also!

  • May 22, 2013 | 11:47 p.m.

    Interesting. I agree about the link between neuroscience and psychology. Can you explain more what you mean about spatial processes being more active when effectors were unknown?

  • Icon for: Jaron Colas

    Jaron Colas

    Presenter
    May 23, 2013 | 12:58 a.m.

    Thank you! Good question.

    For these decisions the options under consideration were neatly mapped onto regions of space. When deciding with the effectors unknown, one can still assign value to and compare regions of space rather than the effectors and actions necessary to reach those regions of space. These spatial processes also drive decisions with well-defined effectors and actions.

    Similar comparison processes operating in the spatial domain are involved in your green frogs’ decisions about optimal habitats. Great work as well!

  • May 23, 2013 | 10:05 a.m.

    Gotcha, cool thanks! Yes if only we could map some frog brain activity! I’m increasingly interested in spatial reasoning and how it influences animal navigation and habitat selection.

  • Further posting is closed as the competition has ended.

Icon for: Jaron Colas

JARON COLAS

California Institute of Technology
Years in Grad School: 2

Decision making in an embodied system: Effector-specific and effector-independent neural computations underlying value-based decision making

Decision making generally translates to action selection because every organism is an embodied system that must decide how to interact with the physical world. Neural comparison processes operate at multiple levels of representation insofar as different neural networks represent options under consideration and actions for obtaining them at different levels. How does the nervous system compute decision values at different levels of representation and integrate them? This research has begun to shed light on effector-specific and effector-independent neural computations underlying value-based decision making with the aim of achieving a complete mechanistic understanding of how people make and implement decisions.