We were able to show that we could replicate verbal overshadowing with the addition of a number of filler tasks.
The purpose of this study is to create a maximum amount of time participants can spend on the filler tasks. WHile the total length of the study was supposed to be 25 minutews, participants spent an average of 45 minutes on the study. This is attributable to spending more than 20 minutes on the filler task and is the most likely reason why the base-rate of correctly identifying the suspect was so much lower in our replication (M = 31%) than in the mTurk RRR (Alogna et al., 2014).
We will thus drop a couple of the filler tasks and include a survey flow element where after each filler task the total duration is checked and if it is equal to or longer than 22 minutes the participant will automatically move on to the randomization, DV, and checks to finish the study.
We will again collect N = 1,096 participants per our previous power analysis.
These 1,096 will be drawn in a stratified way with unequal probabilities of selection, so that the people who complete each survey will resemble the nation's adult population (according to the most recently available Current Population Survey, conducted by the U.S. Census Bureau) in terms of gender, age, education, ethnicity (Hispanic vs. not), race (allowing each respondent to select more than one race), region, and income.
Anyone who was invited to take the first study will not be invited to take this study.
**References**
Alogna, V. K., Attaya, M. K., Aucoin, P., Bahník, Š., Birch, S., Birt, A. R., ... & Buswell, K. (2014). Registered replication report: Schooler and engstler-schooler (1990). Perspectives on Psychological Science, 9(5), 556-578.