Data and code for: A Unified Model of Learning to Forecast

We propose a model of boundedly rational and heterogeneous expectations that unifies adaptive learning, k-level reasoning, and replicator dynamics. Level-0 forecasts evolve over time via adaptive learning. Agents revise over time their depth of reasoning in response to forecast errors, observed and counterfactual. The unified model makes sharp predictions for when and how fast markets converge in Learning-to-Forecast Experiments, including novel predictions for individual and market behavior in response to announced events. <br><br>This repository contains
the raw data of the experiment to the test the unified model. It
contains experimental data collected in the UNSW Sydney Business School BizLab
in 2018 and in the University of Sydney School of Economics Experimental Lab in
2019. Subject played a learning-to-forecast game, where they were paid based on
the accuracy of their forecasts. The subjects forecasted the price of a good traded
in simple demand and supply dynamic market environment with a production lag.
The market featured structural changes to the environment, which were fully
known to subjects. How subjects' forecasts incorporated the information about
the structural change is the main question of interest. <br>