Thinking of learning phenomena as instances of relational behavior.

We explore the idea that some learning phenomena can be thought of as instances of relational behavior-more specifically, arbitrarily applicable relational responding (AARR). After explaining the nature of AARR, we discuss what it means to say that learning phenomena such as evaluative and fear conditioning are instances of AARR. We then list several implications of this perspective for empirical and theoretical research on learning, as well as for how learning phenomena relate to other psychological phenomena in human and nonhuman animals.

Effects on the Affect Misattribution Procedure are strongly moderated by influence awareness.

The Affect Misattribution Procedure (AMP) is used in many areas of psychological science based on the assumption that it not only taps into attitudes and biases but does so without a person's awareness. Across eight preregistered studies (N = 1603) plus meta-analyses, we reexamined the 'implicitness' of AMP effects, and in particular, the idea that people are unaware of the prime's influence on their evaluations. Results indicated that AMP effects and their predictive validity are primarily moderated by a subset of influence-aware trials (within individuals), and high rates of influence awareness (between individuals). Interestingly, an individual's influence-awareness rate on one AMP predicted how they performed on an earlier AMP, even when the two assessed different attitude domains. Taken together, our results suggest that AMP effects are not implicit in the way that has been claimed, a finding that has implications for the procedure, past findings, and theory. All materials and data are available at osf.io/gv7cm.

Assessing sleep-related attitudes with the implicit association test: A prospective study in young adults.

The measurement of automatic attitudes towards sleep, in addition to reflective self-reports, might improve our ability to predict and explain sleep-hindering practices. Two types of implicit association tests (IATs), a sleep-related evaluations IAT and a sleep-related self-identity IAT, were developed to evaluate their efficacy for assessing automatic sleep-related attitudes. In addition, a speeded self-report measure of sleep evaluations was explored as a means to assess automatic sleep-related attitudes. The study included 136 young adults (age = 21.70 ± 2.22, 43% female). At baseline, the two IATs, the speeded self-report, and standard self-reports of sleep determinants (reflective attitudes, self-efficacy, intention and action planning for sleep-promoting behaviour), sleep hygiene practices, sleep quality, and sleep duration were assessed. All variables except for the sleep determinants were assessed again at 2-week follow-up. The results demonstrated good reliability of the two IAT versions, but both IATs were unrelated to the speeded self-report, the sleep determinants, sleep practices, sleep quality or sleep duration. The speeded self-report correlated significantly with the standard self-reports of sleep determinants. Baseline scores on the IATs or speeded self-report did not predict sleep hygiene practices, sleep duration or sleep quality at follow-up. The findings indicate that sleep-related IATs might not be suited to assess automatic sleep-related attitudes. Further investigation is needed to determine whether speeded self-reports are valid measures of automatic attitudes. Moreover, more empirical research is required to clarify the role of automatic processes for sleep hygiene behaviours.

An implicit measure of growth mindset uniquely predicts post-failure learning behavior.

Research on implicit theories of intelligence (a.k.a. intelligence mindset) has shown that endorsing a stronger growth mindset (the belief that intelligence can be improved) is adaptive in the face of difficulties. Although the theory presumes implicit processes (i.e., unaware beliefs, guiding behaviors and actions automatically), the concept is typically assessed with self-reports. In this project we brought together research on intelligence mindset with research on implicit social cognition. Harnessing recent innovations from research on implicit measures, we assessed intelligence mindsets on an implicit level with a mousetracking Propositional Evaluation Paradigm. This measure captures the spontaneous truth evaluation of growth- and fixed-mindset statements to tap into implicit beliefs. In two preregistered laboratory studies (N = 184; N = 193), we found that implicitly measured growth mindsets predicted learning engagement after an experience of failure above and beyond the explicitly measured growth mindset. Our results suggest that implicit and explicit aspects of intelligence mindsets must be differentiated. People might be in a different mindset when making learning-related decisions under optimal conditions (i.e., with ample time and capacity) or under suboptimal conditions (i.e., when time pressure is high). This advancement in the understanding of implicit theories of intelligence is accompanied with substantial implications for theory and practice.

Relating different perspectives on how outcomes of behavior influence behavior.

Many researchers have tackled the question of how behavior is influenced by its outcomes. Some have adopted a nonmechanistic (functional) perspective that attempts to describe the influence of outcomes on behavior. Others have adopted a mechanistic (cognitive) perspective that attempts to explain the influence of outcomes on behavior. Orthogonal to this distinction, some have focused on the influence of outcomes that a behavior had in the past, whereas others also consider the influence of outcomes that a behavior might have in the future. In this article, we relate these different perspectives with the goal of reducing misunderstandings and fostering collaborations between researchers who adopt different perspectives on the common question of how behavior is influenced by its outcomes.

The development of an implicit measure of treatment expectations.

In three preregistered studies, we investigated whether implicit treatment expectations, using a relational implicit measure (the MT-PEP), vary between participants provided opposing information about novel medical treatments (Studies 1 and 2) or who responded based on normative beliefs toward common over-the-counter drugs (Study 3). The studies revealed large Cohen's d effect sizes of both novel and well-known treatment information within the implicit measure. The studies also provide evidence of convergent validity, with MT-PEP scores associated with explicit beliefs about medicine and over-the-counter drug familiarity. Implicit treatment expectations can be assessed and offer a novel tool for the intersection of social psychology and medicine.

How to Build and How not to Build an Implicit Measure in Behavior Analysis: A case Study Using the Function Acquisition Speed Test.

This article provides a comprehensive overview of the development of a behavior-analytic alternative to the popular implicit association test (IAT), namely, the function acquisition speed test (FAST). The IAT appears, prima facia, to indirectly assess participants' learning histories with regard to the categorization of stimuli. However, its origin within cognitive psychology has rendered it replete with mentalism, conceptual ambiguity, statistical arbitrariness, and confounding procedural artifacts. The most popular behavioral alternative to the IAT, the widely used implicit relational assessment procedure (IRAP), has inherited many of these concerning artifacts. In this article, we present a behavior-analytic critique of both the IAT and IRAP, and argue that a behavior-analytic approach to implicit measures must have stimulus control front and center in its analysis. We then outline a series of early research studies that provided the basis for a potentially superior procedure within our field. We go on to outline how this early research was harnessed in stepwise research, guided by a strict adherence to traditional behavior-analytic methods for the analysis of stimulus relations, to increasingly modify a test format fit for the behavior analyst interested in assessing stimulus relatedness.

ABC-training as a new intervention for hazardous alcohol drinking: Two proof-of-principle randomized pilot studies.

BACKGROUND AND AIMS: ABC-training is a new intervention to encourage health behavior change that targets the automatic activation of adaptive beliefs (i.e. automatic inferences). The aim of this proof-of-principle study was to test the effectiveness of web-based ABC-training to change outcome expectancies of alcohol drinking in a sample of hazardous drinkers. DESIGN: One exploratory and one confirmatory experiment with two between-subject conditions (online ABC- and control-training) and assessments at baseline and 1 week later (after three sessions of training). SETTING: Participants recruited on Prolific Academic completed the web-based study. PARTICIPANTS: Adults with self-reported hazardous alcohol drinking (Experiment 1: 193 adults, United Kingdom, age mean = 46.7 years; Experiment 2: 282 adults, different nationalities, age mean = 38.3 years). INTERVENTION AND COMPARATOR: ABC-training involved completing an online task that required choosing personally relevant alternative behaviors to drinking alcohol in personally relevant antecedent contexts to attain personally important outcomes. Comparator was control-training, in which participants selected both the alternative behaviors and alcohol drinking an equal number of times. Training was completed at baseline, after 3 days and after 1 week. MEASUREMENTS: Primary outcome was change in automatic and self-reported (negative/positive) outcome expectancies of alcohol drinking from baseline to after 1 week. Secondary outcomes were change in weekly alcohol consumption, self-efficacy, craving and motivation (and approach-alcohol associations in Experiment 1). Moderators were baseline outcome scores, motivation, age and alcohol dependency. FINDINGS: Findings of this study are as follows: stronger increase in negative outcome expectancies after ABC- than control-training (Experiment 1: self-report, 95% confidence interval of difference scores (CIdiff ) = [0.04, Inf]; automatic, CIdiff = [0.01, Inf]; Experiment 2: self-report, CIdiff = [0.16, Inf]; automatic, CIdiff = [0.002, Inf]). Stronger reduction in self-reported positive outcome expectancies after ABC- than control-training (Experiment 1: CIdiff = [-Inf, -0.01]; Experiment 2: CIdiff = [-Inf, -0.21]) but mixed findings on automatic positive outcome expectancies (Experiment 1: CIdiff = [-Inf, 0.02]; Experiment 2: CIdiff = [-Inf, -0.001]). CONCLUSIONS: ABC-training may change outcome expectancies of alcohol consumption, but testing of clinically relevant effects in other samples is warranted.

Examining Automatic Stereotyping From a Propositional Perspective: Is Automatic Stereotyping Sensitive to Relational and Validity Information?

Research on automatic stereotyping is dominated by the idea that automatic stereotyping reflects the activation of (group-trait) associations. In two preregistered experiments (total N = 391), we tested predictions derived from an alternative perspective that suggests that automatic stereotyping is the result of the activation of propositional representations that, unlike associations, can encode relational information and have truth values. Experiment 1 found that automatic stereotyping is sensitive to the validity of information about pairs of traits and groups. Experiment 2 showed that automatic stereotyping is sensitive to the specific relations (e.g., whether a particular group is more or less friendly than a reference person) between pairs of traits and groups. Interestingly, both experiments found a weaker influence of validity/relational information on automatic stereotyping than on non-automatic stereotyping. We discuss the implications of these findings for research on automatic stereotyping.

Do evaluation and self-identification relate to self-reported and actual social media use?

Background and aims: Social media use can sometimes become excessive and damaging. To deal with this issue, scholars and practitioners have called for the development of measures that predict social media use. The current studies test the utility of evaluation and self-identification measures for predicting social media use. Method: Study 1 examined the relation between evaluation (n = 58) and self-identification (n = 56) measures on the one hand and several self-report measures of social media use on the other hand. Study 2 examined whether the main results of Study 1 could be replicated and whether evaluation (n = 68) and self-identification (n = 48) also relate to actual social media use. We probed evaluation and self-identification using implicit and explicit measures. Results: Explicit evaluation and self-identification measures significantly correlated with several of the self-report measures of social media. Explicit evaluation also significantly correlated with several indices of actual social media use. Implicit measures did not relate to social media use. Discussion and conclusions: The current results suggest that researchers and practitioners could benefit from using explicit evaluation and self-identification measures when predicting social media use, especially an evaluation measure since this measure also seems to relate to actual social media use. Study 2 was one of the first to test the ecological validity of social media use measures. Although implicit measures could provide benefits for predicting social media use, the current studies did not show evidence for their predictive utility.

On the role of (implicit) drinking self-identity in alcohol use and problematic drinking: A comparison of five measures.

Objective: Implicit and explicit drinking self-identity appear to be useful in predicting alcohol-related outcomes. However, there are several different implicit and explicit measures which can be used to assess drinking self-identity. Some of these implicit measures can also capture relational information (e.g., I am a drinker, I should be a drinker), which might provide unique advantages. Despite the importance of having good measures of drinking self-identity, to date there has been little direct comparison of these measures. Method: This study (N = 358) systematically compared two commonly used measures of drinking self-identity (one implicit and one explicit: the Implicit Association Test [IAT] and the Alcohol Self-Concept Scale [ASCS]) with three relational measures of implicit self-identity (the autobiographical IAT [aIAT], the Relational Responding Task [RRT], and the Propositional Concealed Information Test [pCIT]) on a range of criteria relevant to experimental and clinical alcohol researchers. Results: Overall, we found mixed performances on the implicit measures. Interestingly, the aIAT, which probed should-based drinking identity, performed better than the standard IAT. However, the explicit measure exhibited superior performance to all other measures across all criteria. Conclusions: Our results suggest that researchers who wish to assess drinking-related self-identity and to predict alcohol-related outcomes cross-sectionally should set their focus primarily on the use (and further development) of the ASCS, rather than any of the implicit measures. Future research focusing on the ASCS should seek to investigate the generalizability of our findings to patient populations, and incorporate relational information within that procedure to further improve upon its already-strong utility. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Measuring differential nodal distance using the function acquisition speed test.

Increasing evidence suggests that the relatedness of stimuli within the Function Acquisition Speed Test (FAST) methodology is sensitive to the learning histories of participants. For example, this method is sensitive to differences in the amount of baseline training provided to establish stimulus equivalence relations using arbitrary stimuli (Cummins et al., 2018a). However, it has not yet been investigated whether the relatedness of stimuli within the FAST varies based on differential nodal distances between stimuli within stimulus classes. If so, the FAST could serve an important adjunct assessment procedure for researchers who wish not only to assess the formation of stimulus classes using traditional methods, such as matching-to-sample, but also the relative relatedness of stimuli within complex stimulus classes (i.e., nodal distance). The current study sought to investigate this possibility. Participants (n = 16) were trained in the formation of two 4-member equivalence classes consisting of arbitrary nonsense syllables. Following this, participants completed three FAST assessments, each of which probed for the relatedness of stimulus pairs of differing nodal distance. Group- and individual-level analyses broadly demonstrated that relatedness varied as a function of nodal distances in pre-trained stimulus classes. However, results also highlighted some limitations of the FAST at the individual level.

An inkblot for beliefs: The Truth Misattribution Procedure.

An increasing body of evidence shows the importance of accommodating relational information within implicit measures of psychological constructs. Whereas relational variants of the Implicit Association Test (IAT) have been proposed in the past, we put forward the Truth Misattribution Procedure (TMP) as a relational variant of the Affect Misattribution Procedure (AMP) that aims to capture implicit beliefs. Across three experiments, we demonstrate that TMP effects are sensitive to the relational information contained within sentence primes, both in the context of causal stimulus relations of a known truth value (e.g., "smoking causes cancer" vs. "smoking prevents cancer"), as well as in the domain of gender stereotypes (e.g., "men are arrogant" vs. "men should be arrogant"). The potential benefits of the TMP are discussed.

The relationship between differential stimulus relatedness and implicit measure effect sizes.

Implicit measures have been hypothesized to allow researchers to ascertain the existence and strength of relations between stimuli, often in the context of research on attitudes. However, little controlled behavioral research has focused on whether stimulus relations, and the degree of relatedness within such relations, are indexed by implicit measures. The current study examined this issue using a behavior-analytic implicit-style stimulus relation indexing procedure known as the Function Acquisition Speed Test (FAST). Using a matching-to-sample (MTS) procedure to train stimulus equivalence relations between nonsense syllables, the number of iterations of the procedure was varied across groups of participants, hence controlling stimulus relatedness in the resulting equivalence relations. Following final exposure to the MTS procedure, participants completed a FAST. Another group of participants was exposed to a FAST procedure with word pairs of known relatedness. Results showed that increasing relatedness resulted in a linear increase in FAST effect size. These results provide the first direct empirical support for a key process-level assumption of the implicit literature, and offer a behavior-analytic paradigm within which to understand these effects. These results also suggest that the FAST may be a viable procedure for the quantification of emergent stimulus relations in stimulus equivalence training.