**Original citation.** Kraljic, T., Samuel, A.G., & Brennan, S.E. (2008). First Impressions and Last Resorts: How listeners adjust to speaker variability. *Psychological Science, 19*(4), 332-338.
**Target of replication.** We seek to replicate Kraljic et al.'s finding that if there is strong evidence that an initial impression is an incidental example of speech, the primacy bias should be overridden and participants will not learn that the speech example is characteristic. They suggest that participants only learn when there is no direct evidence of incidence—i.e. when the speech is characteristic, F(1, 260) = 5.08, prep= .92.
**A priori replication criteria.** A successful replication would find that the mean amount of learning for characteristic attribution will be greater than for incidental attribution.
To test for the main effect of attribution manipulation, we will run a 2 (attribution: characteristic of the speaker vs. incidental consequence of some temporary state) x 2 (modality: audio only vs. audiovisual presentation) between subjects ANOVA.
All participants will have normal or corrected to normal hearing and vision. All participants additionally will be native English speakers. As in the original study by Kraljic et al., participants who are extremely inaccurate on the lexical-decision task (20% or below) will be excluded from the analyses.
**Data collection is still underway.** It is expected to be completed in spring 2015.
