Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Experimental Design, the Conlisk-Watts Assignment Model, and the Proper Estimation of Behavioral Response