Implicit cognition measures such as the Implicit Association Test (IAT; Greenwald, McGhee, &, Schwartz, 1998) have become increasingly common in many areas including research on prejudice, consumer preferences, political attitudes, psychopathology, and personality traits (Greenwald, Poehlman, Uhlmann, & Banaji, 2009), in part to avoid the problems of self- report such as susceptibility to self -presentation biases and introspective limits (Greenwald & Banaji, 1995). Implicit measures provide important additional information to explicit assessments, particularly in domains heavily affected by social desirability (Greenwald et al., 2009) or with automatic and spontaneous behaviors (e.g., Asendorpf, Banse, & Mucke, 2002). These measures have not become common in clinical settings, however, due to their procedural characteristics. The IAT, currently the most popular method, relies on the finding that individuals are generally faster at sorting stimuli based on two concepts to the same response key when these concepts are associated than when they are not. For example, an individual may be faster at sorting words related to flower and good to the same key than at sorting words related to flower and bad to the same key. There are hundreds of studies on the IAT (Greenwald et al., 2009), but it only assesses the relative strength of target concepts (De Houwer, 2002), which greatly limits its applied use. For example, if the IAT shows faster responding with flower-good/insect-bad trials than flower-bad/insect-good trials, it is unclear whether the effect is due to a flower-good association, insect-bad association or some relative contribution of both. The IAT design also limits its applicability to domains that go beyond simple associations and bipolar categories, which is often the case with the kinds of beliefs and attitudes applied issues present. Researchers have been working on a variety of alternative IAT designs (e.g., Cohen, Beck, Brown, & Najolia, 2010; Karpinski & Steinman, 2006; Nosek & Banaji, 2001), but none yet overcome these problems. Relational Frame Theory (RFT; Hayes, Barnes-Holmes, & Roche, 2001) researchers have developed an alternative measure, the Implicit Relational Assessment Procedure (IRAP; Barnes-Holmes et al., 2006). Participants are asked to select a relation (e.g., similar/different) between a target stimulus (e.g., substance user words) and a label stimulus (i.e., good/bad) in a series of trials. Two types of trial blocks are used, one in which the verbal relations are consistent with the participants' history of relating stimuli (e.g., addict is similar to bad) and the other where the responses are inconsistent (e.g., addict is similar to good). Participants are trained to emit these two opposing types of sorts (i.e., consistent and inconsistent responses) through an alternating series of practice trials. The difference in response latency between consistent and inconsistent trial blocks in subsequent testing is used to detect the implicit effect. The IRAP is more flexible than the IAT, particularly as it can be used to examine specific implicit relations with a target concept, rather than only relative associations, and to assess a broad range of relations beyond associations. The IRAP demonstrates predicted differences between known groups in a wide variety of areas including some of applied relevance such as self-esteem (Scanlon, Barnes-Holmes, Barnes-Holmes, & Stewart, under review; Vahey, Barnes-Holmes, Barnes-Holmes, & Stewart, 2009), attitudes towards different nationalities (Power, Barnes-Holmes, Barnes-Holmes, & Stewart, 2009), and sexual attitudes (Dawson, Barnes-Holmes, Gresswell, Hart, & Gore, 2009) among many others. The IRAP diverges from explicit self- reports in predicted ways (Barnes-Holmes, Murphy, Barnes-Holmes, & Stewart, 2010; Power et al., 2009), and is sensitive to variables that go beyond explicit reports (Roddy, Stewart & Barnes-Holmes, 2010). …
Read full abstract