APS Replication Initiative Under Way

Earlier this year, Perspectives on Psychological Science announced the first project of the new Registered Replication Report (RRR) initiative, which aims to support high-quality, multi-center replications of important psychological findings. The response was tremendous: 30 labs from all over the world are currently participating in our first RRR project, a replication of Jonathan Schooler and Tonya Engstler-Schooler’s 1990 study on the verbal overshadowing effect (Schooler and Engstler-Schooler, 1990). Now, researchers from the Netherlands to New Zealand are replicating the original study, in which participants who watched a video of a fake robbery and then verbally described the robber were found to be less likely to correctly identify him in a subsequent lineup than those who had performed an unrelated cognitive task after “witnessing” the staged crime. The idea is that when we describe observed events, such as when we are interviewed by a police officer as a witness, we create a verbal memory that may leave out some details or include distorted ones, and this verbal memory can interfere with (or overshadow) our visual memory. The current RRR project aims to clarify the size of this effect, an estimate that will be calculated from a broad and truly global sample.

Some of the participating labs have already completed their data collection; the others will be finished by the end of the year. After each lab writes up its results, University of Illinois Professor Daniel Simons, who is editing the RRR series along with Alex Holcombe of the University of Sydney and Perspectives Editor Bobbie Spellman, will compile these individual results in a meta-analysis to estimate the true effect size of the verbal overshadowing effect. The final Registered Replication Report, which will include the meta-analysis and individual lab findings, will be published in Perspectives in the spring of 2014, and the authors of the original study will be invited to submit a commentary. The report will be open access, and each lab’s raw data will be publicly available.

Several more replication projects are in the works, including a replication of a 2011 study published in Psychological Science by William Hart and Dolores Albarracín entitled “Learning About What Others Were Doing: Verb Aspect and Attributions of Mundane and Criminal Intent for Past Actions” (Hart and Albarracín, 2011). This article deals with grammatical aspect, or how an action is represented in time — specifically the difference between imperfective aspect, which indicates an action within the flow of time (“He was going to the store.”), and perfective aspect, which explicitly indicates the action is bounded or closed in time (“He went to the store.”).

Previous research has found that people tend to associate imperfective aspect with ongoing action and perfective aspect with completed action. In other words, when most people hear “he was walking to the kitchen,” they imagine a man in motion, arms swinging as he moves through the house; “he walked to the kitchen,” on the other hand, tends to invoke an image of a man standing in the kitchen, arrived at his destination. Hart and Albarracín found a related difference in the attribution of intentionality: subjects tended to interpret greater intentionality in people whose actions were expressed using imperfective aspect. This has serious implications for courtroom proceedings, as describing a defendant who was shooting a gun versus one who shot a gun could potentially leave very different impressions on judges and jurors.

Because of the significant implications of this finding for the legal system as well as the field of psycholinguistics, researchers such as RRR proposer Anita Eerland are interested in determining a more precise size for this verb aspect effect through a multi-center replication project. Once the protocol is finalized, it will be posted on the APS website and interested researchers can submit a secondary proposal to participate in the replication.

With one project underway, one in the final stages of preparation, and several more in development, the Registered Replication Reports initiative is already taking off, but it’s also just getting started. The editors continue to consider replication proposals on an ongoing basis; more information about submitting a proposal can be found here.