New Research From Psychological Science

Read about the latest research published in Psychological Science:

The Relative Trustworthiness of Inferential Tests of the Indirect Effect in Statistical Mediation Analysis: Does Method Really Matter?

Andrew F. Hayes and Michael Scharkow
Mediation analysis is commonly used to examine the indirect effect of one variable (X) on a second variable (Y) through a mediator variable (M). Although many researchers conduct mediation analysis, not all do so in the same way. The authors examined the delta method, percentile and bias-corrected bootstrap confidence intervals (CIs), the distribution-of-the-product approach, the Monte Carlo CI, and the test of joint significance. Their findings indicate that although the results of two different tests were more likely to agree than to disagree, different statistical methods address different concerns — for example, if the main concern is power, it’s best to use a bias-corrected bootstrap CI.

Two Items Remembered as Precisely as One: How Integral Features Can Improve Visual Working Memory

Gi Yeul Bae and Jonathan I. Flombaum

Past research has found that as the number of items stored in visual working memory (VWM) increases, the precision of the memory for those items decreases. Researchers have speculated that this occurs because the items are competing for limited memory resources. The authors challenged this hypothesis by having participants perform memory tests in which the integral features of objects, such as size and shape, were manipulated. The researchers found a decline in memory when two items shared an integral feature but not when the objects possessed different integral features. This indicates that declines in VWM are due to correspondence challenges — problems determining which memories correspond to each object — rather than to competition for limited memory resources.

Adam D. Galinsky, Cynthia S. Wang, Jennifer A. Whitson, Eric M. Anicich, Kurt Hugenberg, and Galen V. Bodenhausen
Derogatory group labels are usually used to express contempt and derision, but some groups adopt the very slurs others use against them. The researchers conducted 10 studies in which participants reported on the effects of derogatory labels that were self-imposed or imposed by others. The findings supported a model proposed by the researchers in which self-labeling with a stigmatizing term increases both individual and group perceptions of power, thereby reducing the negative impact of the label. 

Adaptive Memory: The Mnemonic Value of Animacy

James S. Nairne, Joshua E. VanArsdall, Josefa N. S. Pandeirada, Mindi Cogdill, and James M. LeBreton

Research has suggested that humans evolved to prioritize animate (i.e., living) objects because it maximizes the change of detecting potential predators. If this is true, it would make sense for humans to prioritize processing and remembering of animate things. To test this, the researchers had participants memorize and recall a list of 24 animate (e.g., baby, bee) and inanimate (e.g., hat, drum) objects. Participants correctly recalled more animate words than inanimate words, which provides evidence for an evolutionarily based processing advantage for animate words.

Warnings of Adverse Side Effects Can Backfire Over Time

Yael Steinhart, Ziv Carmon, and Yaacov Trope

Common sense dictates that warnings on products — such as cigarettes and medications — make those products less desirable. In one of several studies, participants saw an advertisement for cigarettes that did or did not include a health warning and were told they could buy the cigarettes and receive them within 24 hours or in 3 months. Those who saw the ad with the warning bought fewer cigarettes if they would receive them in 24 hours and more cigarettes if they would receive them in 3 months. This indicates that product warnings might actually backfire in situations in which there is a long interval between the warning and the related behavior.