It is worth noting, however, that the “warm-cold” effect reported here was in reality a cold effect. While an imagined iced coffee reduced estimates of warmth, ratings of the target did not differ reliably from baseline in the hot-coffee condition. This evaluative asymmetry mirrors previous findings (e.g., Bargh & Shalev, 2012) and corroborates the observation that positivity (i.e., warmth) is the default state in person appraisal (Cacioppo & Gardner, 1999).