PUB - Publikationen an der Universität Bielefeld

**Abstract**
Intolerance of uncertainty (IU) is a transdiagnostic risk factor for internalizing disorders. Prior work has found that IU may be associated with either increased reactivity to threat or, alternatively, with decreased differential responding between threat and nonthreat/safety cues (i.e., threat generalization). For example, work by Morriss, Macdonald, & van Reekum  (2016) found that higher IU was associated with increased threat generalization during acquisition (using skin conductance response (SCR)), as well as less differentiation between acquisition and extinction (using subjective uneasiness ratings). Here, three labs attempted direct and conceptual replications of Morriss, Macdonald, et al. (2016). Results showed that the direct replication failed, despite being conducted at the same lab site as the original study; moreover, in contrast to Morriss, Macdonald, et al. (2016), the direct replication found that higher IU was associated with greater SCRdiscriminationbetween threat and safety cues (across acquisition and extinction), as well as greater differences in uneasiness ratings between acquisition and extinction. Nonetheless, in the conceptual replications, higher IU was associated with greater threatgeneralization, as well as less discrimination between acquisition and extinction, as measured using SCR. Higher IU was also associated with larger late positive potentials to threat versus safety cues during extinction—results that mirror those observed by Morriss, Macdonald, et al. (2016) using SCR. Results are discussed with regards to the challenge involved in defining a successful replication attempt, the benefits of collaborative replication and the use and reliability of multiple measures.

---

Both direct and conceptual replications are essential in validating prior findings. In prior work (Morriss, Macdonald, & van Reekum, 2016), greater IU was associated with increased threat generalization, in which nonthreatening cues that resembled threat‐conditioned stimuli elicited greater than expected SCR. Here, three lab sites (including the original lab site) attempted replications of this work, with extension to event‐related potentials. The direct replication failed to find evidence of an association between IU and threat generalization; however, evidence of greater threat generalization among participants with higher IU was found in the conceptual replications. Results are discussed with regards to the challenge involved in defining a successful replication attempt and the benefits of collaborative, multimethod replication and extension.