Replication in Psychology: Approaches, Perceived Threat, Attributions, and Attitudes Regarding Replication Research and Outcomes Among Social Psychologists

Psychological science is in a state of self-assessment, wherein an expansion of meta-research to examine the reproducibility of psychological findings has emerged. The primary method of assessing the reproducibility of psychological science has been large scale direct replication attempts. Initial replicable estimations are weak, leading some to declare that psychological science is in a crisis of replication. The purpose of this dissertation is to (1) establish the normative research practices of a sample of psychological scientists, (2) examine how threat is associated with different approaches to replication, (3) compare the causes that are attributed to partial replication outcomes for oneself versus others, and (4) explore how various factors (e.g. normative practices, attributions) are related to attitudes toward replication research in psychology. A sample of primarily American social psychologists were surveyed (144 complete cases). Prevailing approaches to replication were, on average, viewed favorably, and contrary to expectations respondents were unthreatened by two replication request scenarios. Respondents assigned dispositional attributes (e.g., flexible research practices) more so as contributing to a partial replication outcome of a peer’s work than their own work. A consistent finding emerged wherein more favorable attitudes toward and support for replication research were associated with greater assignment of flexible research practices on the part of others, and greater assumed selective reporting of studies that “work” by others. Combined, these findings suggests that at the individual-level threat is not a formidable barrier thwarting replication, and associating a partial replication outcome with increased flexible practices on the part of a peer fosters support and favorable attitudes toward replication