Report summary

Harvard University Department of Biostatistics in 2015-2017 reads as a 50-PI network with 67 internal PI collaborations. The research center of gravity is Medicine as the leading field (30% of slots across 28 PIs; 28 labels), Statistics and Probability as the leading subfield (12% of slots across 14 PIs; 14 labels), and Advanced Causal Inference Techniques as the leading topic (6% of slots across 8 PIs; 8 labels). That is promising: collaboration is not just one large hairball, and the leading breakdown groups cover 20 PIs across distinct clusters. The most prominent PIs by weighted works are Walter C. Willett (22.1 weighted works; Nutritional Studies and Diet, Diet and metabolism studies); Tyler J. VanderWeele (21.5 weighted works; Religion, Spirituality, and Psychology, Advanced Causal Inference Techniques). The strongest pairings are Walter C. Willett and A. Heather Eliassen (28 shared works, weight 7.3); Bernard Rosner and A. Heather Eliassen (22 shared works, weight 7.2). The strongest breakdown groups are group 1 with 10 PIs, 14 internal connections, weight 38.6, around Statistics and Probability, Economics and Econometrics, Oncology, led by Walter C. Willett, Tyler J. VanderWeele, Bernard Rosner; group 2 with 7 PIs, 10 internal connections, weight 15.4, around Cancer Research, Genetics, Statistics and Probability, led by Francesca Dominici, Patrick Y. Wen, Michael J. Birrer.