“…Whereas analysts' work may appear anodyne in schemaconsistent settings, they nevertheless respond with more penetrating research when actual 5 performance deviates from expectations generally, and in particular when performance is worse than predicted. Second, we contribute to research on earnings conference calls (Hollander et al, 2010;Matsumoto et al, 2011;Larcker and Zakolyukina, 2012;Mayew and Venkatachalam, 2012;Brochet et al, 2015;Bushee et al, 2015;Davis et al, 2015;Lee, 2016). 2 Direct evidence on the properties of analysts' questions is sparse and limited to assessments of their pooled predictive ability and response to management tone (Allee and DeAngelis, 2015;Brockman et al, 2015;Chen et al, 2015;Milian and Smith, 2015).…”