We introduce the Open Affective Standardized Image Set (OASIS), an open-access online stimulus set containing 900 color images depicting a broad spectrum of themes, including humans, animals, objects, and scenes, along with normative ratings on two affective dimensions-valence (i.e., the degree of positive or negative affective response that the image evokes) and arousal (i.e., the intensity of the affective response that the image evokes). The OASIS images were collected from online sources, and valence and arousal ratings were obtained in an online study (total N = 822). The valence and arousal ratings covered much of the circumplex space and were highly reliable and consistent across gender groups. OASIS has four advantages: (a) the stimulus set contains a large number of images in four categories; (b) the data were collected in 2015, and thus OASIS features more current images and reflects more current ratings of valence and arousal than do existing stimulus sets; (c) the OASIS database affords users the ability to interactively explore images by category and ratings; and, most critically, (d) OASIS allows for free use of the images in online and offline research studies, as they are not subject to the copyright restrictions that apply to the International Affective Picture System. The OASIS images, along with normative valence and arousal ratings, are available for download from www.benedekkurdi.com/#oasis or https://db.tt/yYTZYCga.

Using data from 217 research reports (N = 36,071, compared to 3,471 and 5,433 in previous meta-analyses), this meta-analysis investigated the conceptual and methodological conditions under which Implicit Association Tests (IATs) measuring attitudes, stereotypes, and identity correlate with criterion measures of intergroup behavior. We found significant implicit-criterion correlations (ICCs) and explicit-criterion correlations (ECCs), with unique contributions of implicit (β = .14) and explicit measures (β = .11) revealed by structural equation modeling. ICCs were found to be highly heterogeneous, making moderator analyses necessary. Basic study features or conceptual variables did not account for any heterogeneity: Unlike explicit measures, implicit measures predicted for all target groups and types of behavior, and implicit, but not explicit, measures were equally associated with behaviors varying in controllability and conscious awareness. However, ICCs differed greatly by methodological features: Studies with a declared focus on ICCs, standard IATs rather than variants, high-polarity attributes, behaviors measured in a relative (two categories present) rather than absolute manner (single category present), and high implicit-criterion correspondence (k = 13) produced a mean ICC of r = .37. Studies scoring low on these variables (k = 6) produced an ICC of r = .02. Examination of methodological properties-a novelty of this meta-analysis-revealed that most studies were vastly underpowered and analytic strategies regularly ignored measurement error. Recommendations, along with online applications for calculating statistical power and internal consistency (http://www.benedekkurdi.com/#iat), are provided to improve future studies on the implicitcriterion relationship. Open materials are available under https://osf.io/47xw8/.

Six experiments, involving a total of 6,492 participants, were conducted to investigate the relative effectiveness of repeated evaluative pairings (REP; exposure to category members paired with pleasant or unpleasant images), evaluative statements (ES; verbally signaling upcoming pairings without actual exposure), and their combination (ES + REP) in shifting implicit social and nonsocial attitudes. Learning modality (REP, ES, and ES + REP) was varied between participants and implicit attitudes were assessed using an Implicit Association Test (IAT). Study 1 (N = 675) used fictitious social groups (NIFFs and LAAPs), Study 2 (N = 1,034) used novel social groups (humans with long vs. square faces), Study 3 (N = 1,072) used nonsocial stimuli (squares vs. rectangles), and Study 4 (N = 848) and Study 5 (N = 958) used known social groups (young vs. elderly; American vs. foreign). ES were more effective than REP and no less superior than ES + REP in producing implicit attitude change. Results were robust across social and nonsocial domains and for known and novel groups. Study 6 (N = 1,905) eliminated time on intervention, levels of construal, and expectancy effects as possible explanations for these findings. Associative theories of implicit evaluation posit that implicit attitudes should shift piecemeal over time; yet, in these experiments, one-shot language-based learning led to larger shifts in implicit attitude than exposure to stimulus pairings. Moreover, the redundancy observed in REP + ES suggests that attitude acquisition from repeated pairings and evaluative instructions may rely on shared mental representations. (PsycINFO Database Record

Stereotypes are associations between social groups and semantic attributes that are widely shared within societies. The spoken and written language of a society affords a unique way to measure the magnitude and prevalence of these widely shared collective representations. Here, we used word embeddings to systematically quantify gender stereotypes in language corpora that are unprecedented in size (65+ million words) and scope (child and adult conversations, books, movies, TV). Across corpora, gender stereotypes emerged consistently and robustly for both theoretically selected stereotypes (e.g., work–home) and comprehensive lists of more than 600 personality traits and more than 300 occupations. Despite underlying differences across language corpora (e.g., time periods, formats, age groups), results revealed the pervasiveness of gender stereotypes in every corpus. Using gender stereotypes as the focal issue, we unite 19th-century theories of collective representations and 21st-century evidence on implicit social cognition to understand the subtle yet persistent presence of collective representations in language.