The study by Terbeck et al. (2012) that was published in Psychopharmacology, Online First concluded that propranolol reduces implicit negative racial bias. However, this conclusion is not supported by the study because the implicit association test (IAT) (Greenwald et al. 2009) has not been sufficiently validated and because the claim for clinical treatment due to amygdala suppression is unsupported. The classic race IAT compares whether one is quicker to link European–Americans with words associated with the concept “good” and African–Americans with words related to “bad” or vice versa. Gladwell (2005) summarizes it in the following way: The IAT measures a person’s attitude on an unconscious level or the immediate and automatic associations that occur even before a person has time to think. Is it possible to reliably measure “unconscious attitude” without using presumptions or indirect measures that rely on subjective notions? Blanton et al. (2007) also question the arbitrary nature of the cutoffs for mild, moderate, and strong preferences when no research shows where these limits should be. There is no research to support differences in individuals above or below the cutoffs. The IAT essentially attempts to measure the immeasurable and has been used prematurely in research without sufficient methodological validation. The measurement reliability of the IAT, as determined by test–retest methods, shows negligible consistency. The study only used data from blocks containing less than 15 % errors, but does not state how many of these blocks were discounted, which could be substantial. Moreover, it is mentioned that one participant was excluded due to high errors on both trials, and three participants had onlymain trials analyzed due to extreme errors in the practice session. These breakdowns in measurement quality are further compounded on the IAT developer's website (Greenwald et al. 2012) which states that “besides normal variation in the reliability of assessment, the IAT is also known to be malleable based on differences in the social setting and recent experience. These factors will influence the consistency of measurement across occasions.” The IATcould not have high measurement quality of attitude if so many extraneous factors have such influence. Lastly, the conclusion that the minimal effect seen in response latencies is the result of direct inhibition of amygdala stimulation is not supported. The limitations of previous studies that correlated the participant's IATscore with the extent of amygdala activation are noted, but the same limitations are not applied to the present study. For example, contradictory findings are referenced: “Phelps et al. (2003) found that a patient with bilateral amygdala damage exhibited an intact IATeffect, leading the authors to suggest that the amygdala may not be a critical structure for the manifestation of implicit bias.” In addition, the study concluded that more research needs to be conducted in order to determine whether propranolol acts peripherally or centrally. This means that propranolol may not even directly affect the brain, let alone a specific anatomic structure such as the amygdala. As a last possibility, an explanation that cannot be excluded is that propranolol simply equalizes mental reflex latencies, regardless of task. The study has many strengths, including an excellent review of previous studies, the use of visual analog scales, the use of a control group and two different post-intervention time points, excellent statistical analysis, well-constructed data tables, and most of all, an important topic. However, due to important T. S. Burchett : L. L. Glenn College of Nursing and Institute for Quantitative Biology, East Tennessee State University, P.O. Box 70658, Johnson City, TN 37604, USA
Read full abstract