Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
I. Introduction, 705. — II. The model, 707. — III. Empirical results, 708. — IV. Conclusion, 710.