Chapter 8 Further Reading: Reward Prediction Error and Anticipation

Academic Papers and Research

1. Schultz, W., Dayan, P., & Montague, P.R. (1997). "A Neural Substrate of Prediction and Reward." Science, 275(5306), 1593–1599. The landmark paper establishing the reward prediction error framework. Schultz, Dayan, and Montague jointly presented the neural and computational accounts of dopamine as a prediction error signal, connecting Schultz's empirical recordings from monkey dopamine neurons with the temporal difference learning framework from computer science. Reading the original paper, or at minimum its introduction and conclusions, is essential for understanding the RPE concept in its proper scientific context. The paper is technically dense in parts but the conceptual sections are accessible.

2. Montague, P.R., Dayan, P., & Sejnowski, T.J. (1996). "A Framework for Mesencephalic Dopamine Systems Based on Predictive Hebbian Learning." Journal of Neuroscience, 16(5), 1936–1947. The companion theoretical paper to Schultz et al. (1997), developing the computational account of dopamine as a temporal difference learning signal. Montague, Dayan, and Sejnowski's framework established that the brain's dopamine system implements a biologically feasible version of the TD-learning algorithm from reinforcement learning research — one of the most productive convergences between AI theory and neuroscience. Provides the computational foundation for understanding how the RPE signal drives learning and behavioral change.

3. Kahneman, D., & Tversky, A. (1979). "Prospect Theory: An Analysis of Decision Under Risk." Econometrica, 47(2), 263–292. The foundational paper establishing prospect theory and the psychological asymmetry between losses and gains — the loss aversion principle. While not directly about neuroscience or social media, this paper provides the theoretical foundation for understanding why streak mechanics and other loss-contingent social media features are so behaviorally powerful. The paper's framing of how people actually make decisions under uncertainty, as opposed to how rational agent models predict they will, is essential background for the analysis of social media behavioral mechanics.

4. Rangel, A., Camerer, C., & Montague, P.R. (2008). "A framework for studying the neurobiology of value-based decision making." Nature Reviews Neuroscience, 9(7), 545–556. An influential review paper from the emerging field of neuroeconomics, integrating the RPE framework with broader understanding of how the brain assigns motivational value to stimuli and makes decisions under uncertainty. Rangel, Camerer, and Montague's framework explains how conditioned anticipatory states can distort value computation — making social media checking feel more valuable than it is — through the same mechanisms that produce other forms of motivated cognition. A key bridge paper between neuroscience and behavioral economics.

5. Mark, G., Gudith, D., & Klocke, U. (2008). "The cost of interrupted work: More speed and stress." Proceedings of the ACM Conference on Human Factors in Computing Systems (CHI 2008), 107–110. Gloria Mark and colleagues' influential study documenting that workers interrupted by notifications complete their tasks faster (due to time pressure awareness) but experience significantly higher stress. A companion to the twenty-three minute attention restoration finding, this paper establishes the physiological cost of notification-rich work environments. Essential reading for understanding why the continuous partial attention state produced by social media notifications is not psychologically neutral.

6. Kushlev, K., & Dunn, E.W. (2015). "Checking email less frequently reduces stress." Computers in Human Behavior, 43, 220–228. An experimental study demonstrating that constraining email checking to specific times of day, rather than allowing continuous notification-triggered checking, reduces perceived stress and improves mood. The study's key methodological contribution is the random assignment to checking frequency conditions, allowing causal inference. Its results support the structural intervention approach to managing notification-checking behavior.

7. Haynes, J.D., et al. (2007). "Reading Hidden Intentions in the Human Brain." Current Biology, 17(4), 323–328. Haynes and colleagues' demonstration that brain activity predictive of decisions can be detected up to ten seconds before subjects report awareness of their decision. While not directly about social media, this research has significant implications for the question of voluntary control over checking behavior: if the neural antecedents of the impulse to check precede conscious awareness of the impulse by seconds, the "choice" to check is not straightforwardly voluntary in the sense that everyday intuition suggests.

Books

8. Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux. Daniel Kahneman's comprehensive and accessible synthesis of prospect theory, loss aversion, the distinction between System 1 (fast, automatic, intuitive) and System 2 (slow, deliberate, rational) processing, and decades of behavioral economics research. The chapters on loss aversion are directly relevant to understanding Snapchat streak mechanics and other social media features that exploit the asymmetry between loss and gain. Kahneman's framing of bounded rationality helps explain why individually "irrational" social media behaviors are predictable outputs of well-documented psychological processes.

9. Montague, R. (2006). Why Choose This Book? How We Make Decisions. Dutton Books. Read Montague's accessible account of his computational neuroscience research on decision-making, reward prediction, and the neurobiology of choice. Montague explains the dopamine RPE framework in terms accessible to non-specialists, with extensive discussion of how the same computational principles that govern learning in animals govern human decision-making in social and economic contexts. More accessible than his technical papers and provides excellent entry into the computational psychiatry framework.

10. Alter, A. (2017). Irresistible: The Rise of Addictive Technology and the Business of Keeping Us Hooked. Penguin Press. While discussed in Chapter 7's further reading, Alter's book is particularly relevant to Chapter 8's RPE framework in its treatment of the checking loop and habituation. Alter explicitly discusses the checking behavior loop and its relationship to anticipatory states, and his analysis of behavioral addiction criteria provides a framework for evaluating whether social media checking meets addiction thresholds that is grounded in behavioral science rather than clinical intuition.

11. Sapolsky, R.M. (2011). "Dopamine Jackpot! Sapolsky on the Science of Pleasure." FORA.tv lecture, subsequently widely reproduced. Sapolsky's lecture on dopamine and reward prediction error is one of the most entertaining and accessible introductions to the wanting/liking distinction and the RPE framework available for a general audience. Sapolsky is an extraordinary science communicator, and his account of how the anticipatory dopamine response is stronger for unpredictable rewards than for certain rewards provides an ideal entry point for students encountering these concepts for the first time. Available on YouTube and various educational platforms.

12. Szalavitz, M. (2016). Unbroken Brain: A Revolutionary New Way of Understanding Addiction. St. Martin's Press. Science journalist Maia Szalavitz's account of addiction as a learning disorder rather than a moral failure or a brain disease in the traditional sense. Szalavitz draws on RPE research and learning theory to argue that addiction is fundamentally about misdirected learning — the same neural machinery that enables adaptive learning becomes misdirected toward maladaptive behavioral patterns. The framework she develops is directly applicable to understanding social media compulsive use and provides a compassionate but rigorous alternative to both moralizing and pathologizing accounts of compulsive behavior.

Journalism and Investigative Reporting

13. Bilton, N. (2014, August 26). "Steve Jobs Was a Low-Tech Parent." The New York Times. One of the first mainstream media pieces to document the pattern of technology executives limiting their children's technology use. Jobs described to Bilton restricting his children's use of iPads; subsequent reporting identified similar restrictions among other executives. The contrast between the technology that executives promoted publicly and restricted privately is a recurring theme in the social media accountability literature and suggests private knowledge of behavioral effects not disclosed to the public.

14. Eyal, N. (2014). "The Psychology Behind Instagram." Nir and Far blog. Eyal's early (pre-"Indistractable" reconsideration) analysis of Instagram's engagement mechanics, written from the product design perspective. Eyal applies the Hook Model framework — including RPE dynamics — to Instagram's design, providing a candid inside-out account of how behavioral science was applied to social media design. Reading this piece alongside Eyal's later, more critical work illuminates the evolution of the tech industry's self-understanding on these questions.

15. Twenge, J.M. (2017, September). "Have Smartphones Destroyed a Generation?" The Atlantic. Jean Twenge's accessible summary of her generational data research on adolescent mental health, making the case that the marked decline in adolescent well-being that began around 2012 correlates with the widespread adoption of smartphones and social media. While the causal direction of this correlation remains contested in the academic literature, Twenge's article provides an important data context for understanding the real-world consequences of the behavioral dynamics described in this chapter, particularly for adolescent users.

16. Lorenz, T. (2019, May 13). "Teens Are Being Burdened by Streaks." The Atlantic. Taylor Lorenz's investigation into Snapchat streak dynamics among American teenagers, based on interviews with teens, parents, and school counselors. The reporting provides firsthand qualitative evidence of the streak anxiety, delegated maintenance behavior, and social obligation dynamics described in Case Study 2. Essential reading for grounding the theoretical analysis in the lived experience of the users most directly affected.

Documentaries and Online Resources

17. Schultz, W. (2016). "Dopamine Reward Prediction Error Coding." Dialogues in Clinical Neuroscience, 18(1), 23–32. A relatively accessible review paper by Schultz summarizing decades of RPE research, written for a clinical neuroscience audience rather than a specialist computational neuroscience one. Provides a more recent and complete account of RPE research than the landmark 1997 paper, including discussion of how RPE relates to addiction, decision-making disorders, and human behavior. Available open-access.

18. Harris, T. (2016). "How Technology Hijacks People's Minds — from a Magician and Google's Design Ethicist." Medium: Thrive Global. Former Google design ethicist Tristan Harris's influential essay on the design techniques used by technology platforms to capture and hold attention. Harris's "race to the bottom of the brainstem" argument — that competitive pressure for attention drives platforms to exploit progressively more primitive and less rational neural mechanisms — is a useful complement to the RPE framework developed in this chapter. Harris has since developed this analysis further through the Center for Humane Technology and the documentary "The Social Dilemma."

19. Alter, A. (2017). "Why Our Screens Make Us Less Happy." TED Talk, TED2017. Adam Alter's TED talk on behavioral addiction and screen-based technology is an accessible twenty-minute introduction to the behavioral science of technology compulsion. Alter summarizes his research on the changing patterns of technology use, the criteria for behavioral addiction, and the design features of platforms that meet those criteria. A useful supplement to his book for students who prefer video-format introductions to a topic.

20. Duhigg, C. (2012). The Power of Habit: Why We Do What We Do in Life and Business. Random House. While not focused on social media or neuroscience per se, Duhigg's account of habit formation and change provides a useful framework for thinking about how RPE-conditioned checking behaviors can be modified. Duhigg's "habit loop" (cue, routine, reward) maps directly onto the checking behavior loop, and his analysis of how habits can be modified — by identifying the cue, changing the routine, and maintaining the reward — provides practical complement to the neuroscientific account in this chapter. His treatment of organizational habit change is relevant to the structural intervention argument.