Diverse Perspectives on Strengthening Science

During the last 5 years, Perspectives on Psychological Science (PPS) has published several special sections focused on improving research methods and strengthening psychological science. This special section, introduced by Editor Barbara A. Spellman, is an eclectic collection of articles that present new insights into the strength of our science or provide commentary on topics raised in past methodology-based special sections.

Why are some scientists tempted to use questionable research practices, fabricate data, or otherwise cheat in the pursuit of science? Engel explains this temptation — and how we can reduce it — using a standard economic model describing both the individual benefits to an investigator that can arise from scientific disintegrity and the potential costs to the investigator and to other scientists.

***

Not all types of replications are the same. Exact replications seek to directly repeat a past study, whereas critical replications seek to test alternative explanations of past findings. Larzelere, Cox, and Swindle argue that critical replications are especially important when replicating nonrandomized studies, showing how they can prevent the carryover of systematic biases from previous studies.

***

When conducting a replication, researchers often understand the importance of using a new sample of participants; however, they do not often recognize the importance of applying the same criteria to the stimuli used in the replication. Westfall, Judd, and Kenny discuss the influence of stimuli on the statistical power of a replication and the importance of using new or expanded stimulus sets to protect against the possibility of stimulus-specific effects.

***

In 2014, the first Registered Replication Report confirmed the much studied verbal overshadowing effect. Although researchers may be tempted to provide law enforcement with recommendations based on these results, Mickes and Wixted* argue that this is premature. They say that without knowing the false ID rate of participants — something not studied in the replication — it is difficult to know how verbal descriptions influence the probative value of participants’ identifications.

*John Wixted will speak in the invited symposium “Threshold Recognition vs. Continuous Recognition in the Real World” at the 27th APS Annual Convention in New York, NY, USA.

***

In a 2014 article, Klatzky and Creswell suggested that priming effects may arise from shared perceptual pathways, associative chains, and inferential processes. If the influence of these pathways, chains, and processes is stochastic, then it is not surprising that priming effects are often not replicable. Schwartz suggests that social priming may be an example of a central system operation and explains how these types of phenomena should change the way we view replication failures.

***

Locke* argues that the field of behavioral priming is in need of theory building. Although there is a strong bias for deductive theory building within our discipline, he suggests that the use of inductive theory-building techniques may prove more fruitful. He provides several guidelines that can be used to create inductive theories and uses these guidelines to describe priming theory as it now stands.

*Edwin Locke will speak in the invited symposium “The Illusion of Determinism” at the 27th APS Annual Convention in New York, NY, USA.