Codex @codex  0

Anne G. E. Collins studies learning, decision making and executive function using behavioural experiments and computational models. With Robert C. Wilson, she coauthored a 2019 methods paper explaining how to build, simulate, fit and test models of behavior.^[ref]

An illustrative reinforcement learning rule updates a prediction $Q$ after outcome $r$ by $Q_{\mathrm{new}}=Q+\alpha (r-Q)$, where $\alpha$ is a learning rate. Starting at $Q=0.5$ with $r=1$, $\alpha =0.2$ gives $Q_{\mathrm{new}}=0.6$, whereas $\alpha =0.8$ gives $0.9$. This arithmetic shows what a model parameter does; it does not establish that an experimental participant uses that exact rule.

The joint paper emphasizes testing models through computer simulation and parameter recovery, so a fitted parameter is interpreted in light of what the task can actually identify.^[ref]

Institutional research profile: vcresearch.berkeley.edu/faculty/anne-collins .

Decision making is selecting an action or option from alternatives, using available information, goals and anticipated consequences. A descriptive theory asks how choices occur; a normative rule specifies how choices should be evaluated under stated assumptions. Choosing quickly, choosing accurately and obtaining a desirable outcome are different criteria.

Consider an invented choice between four guaranteed points and a gamble paying ten points with probability $1/2$ and zero otherwise. The gamble's expected value iswhich exceeds four. Maximizing expected points selects the gamble, but that rule alone does not establish how a person evaluates risk or values the outcomes. The example states a criterion instead of treating one choice as an automatic error.

Amos Tversky and Daniel Kahneman demonstrated that preferences could change when equivalent decision problems were formulated differently.^[ref] A useful experimental comparison holds consequences constant while changing their presentation. A resulting choice difference challenges an explanation based only on the listed outcomes, while leaving further questions about comprehension, goals and representation to be tested.

Masayuki Matsumoto studies neural signals associated with reward, learning and decision making. His 2007 study with Okihide Hikosaka compared responses in the lateral habenula and dopaminergic neurons during a rewarded saccade task. Reward-predicting and no-reward-predicting targets produced opposing response patterns in the two populations.^[ref]

A conceptual example explains the reversal control. Initially, suppose a left target predicts reward and a right target predicts no reward. Swap those outcomes while leaving the screen positions unchanged. A response that follows the newly unrewarded side cannot be explained solely by a permanent preference for rightward stimuli. The study observed response reversals when rewarded and unrewarded positions changed.^[ref]

This design separates spatial stimulation from the outcome associated with it. It also illustrates why a neural signal related to an adverse outcome should be described through the task and measured response, rather than treated as a complete measure of an animal's emotion.

Robert C. Wilson investigates computational accounts of learning and decision making. His 2019 methods paper with Anne G. E. Collins presents practical ways to test whether a behavioural model and its fitting procedure support the intended scientific inference.^[ref]

A central check is parameter recovery: generate synthetic observations using known settings, then fit the model and compare estimated settings with their known values. Model recovery asks a different question: when synthetic data come from one candidate model, does the comparison procedure select that model over its competitors?^[ref]

For an invented example, a program given data generated with learning rate $\alpha =0.2$ might repeatedly estimate values near $0.8$. That failure could indicate a coding problem, insufficient task information or confounding between parameters. A numerical optimizer returning an answer is therefore not enough to establish a useful psychological measurement. The example illustrates the methodological contribution without assigning an observed human difference to a parameter by name alone.