Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Introduction, 629.—I. The basic model, 630.—II. Robustness, 638—III. Conclusion, 648.