Fractal Scaling and Implicit Bias: A Conceptual Replication of Correll (2008)

A racial priming article claimed that, relative to a control condition, an exotic variety of variability, called 1/ƒ noise, is altered when stereotypes impact participants’ judgments in an implicit prejudice task (Correll, 2008). However, Madurski and LeBel (2014) recently described two powerful, faithfully cloned, and apparently decisive studies that each failed to return a successful literal replication of Correll’s report. Madurski and LeBel outlined and subsequently eliminated several potential extraneous reasons for their replication failures, such as different participant demographics, participant non-compliance, poor psychometrics, and hardware discrepancies. By contrast, this article reports a successful conceptual replication of the pattern reported by Correll (cf. Schmidt, 2009). Notably, this conceptual replication required adjustments to Correll’s original method and statistical analyses. All the changes were dictated by a systems theory of 1/ƒ noise that was largely in place prior to Correll’s report (Kello, Beltz, Holden, & Van Orden, 2007; Van Orden, Holden, & Turvey, 2003; 2005). Implications for the replication debate are discussed, with emphasis on contextualizing implicit cues.