Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Analyses using aggregated data may bias inference. In this work we show how to aggregate data to avoid or at least reduce this bias when estimating quantile regressions.