Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We report 28 new experiment sessions consisting of up to three experience levels to examine the robustness of learning and "error" elimination among participants in a laboratory asset market and its effect on price bubbles.