“…In particular, effect sizes differed as a function of evaluation potential of the support provider (i.e., low vs. high), mode of support (i.e., silent vs. active), and type of task (i.e., speech vs. math). These findings, along with subsequent work in the area highlight the importance of the support context when assessing links between received support and laboratory stress reactivity (e.g., Craig & Deichert, 2002; Holt-Lunstad, Uchino, Smith, & Hicks, 2007; O'Donovan & Hughes, 2008; Phillips, Gallagher, & Carroll, 2009). …”