With the wide availability of large language models and generative AI, there are four primary paradigms for human–AI collaboration: human-only, AI-only (ChatGPT-4), augmented human (where a human makes the final decision with AI output as a reference), or augmented AI (where the AI makes the final decision with human output as a reference). In partnership with one of the world’s leading consulting firms, we enlisted professional content creators and ChatGPT-4 to create advertising content for products and persuasive content for campaigns following the aforementioned paradigms. First, we find that, contrary to the expectations of some of the existing algorithm aversion literature on conventional predictive AI, the content generated by generative AI and augmented AI is perceived as of higher quality than that produced by human experts and augmented human experts. Second, revealing the source of content production reduces—but does not reverse—the perceived quality gap between human- and AI-generated content. This bias in evaluation is predominantly driven by human favoritism rather than AI aversion: Knowing that the same content is created by a human expert increases its (reported) perceived quality, but knowing that AI is involved in the creation process does not affect its perceived quality. Further analysis suggests this bias is not due to a ‘quality prime’ as knowing the content they are about to evaluate comes from competent creators (e.g., industry professionals and state-of-the-art AI) without knowing exactly that the creator of each piece of content does not increase participants’ perceived quality.

Literature highlights the distinction between intuitive and analytic thinking as a prominent cognitive style distinction, leading to the proposal of various theories within the framework of the dual process model. However, it remains unclear whether individuals differ in their thinking styles along a single dimension, from intuitive to analytic, or if other dimensions are at play. Moreover, the presence of numerous thinking style measures, employing different terminology but conceptually overlapping, leads to confusion. To address these complexities, Newton et al. suggested the idea that individuals vary across multiple dimensions of intuitive–analytic thinking styles and distinguished thinking styles between 4 distinct types: Actively open-minded thinking, close-minded thinking, preference for effortful thinking, and preference for intuitive thinking. They proposed a new measure for this 4-factor disposition, The 4-Component Thinking Styles Questionnaire (4-CTSQ), to comprehensively capture the psychological outcomes related to thinking styles; however, no independent test exists. In the current pre-registered studies, we test the validity of 4-CTSQ for the first time beyond the original study and examine the association of the proposed measure with various factors, including morality, conspiracy beliefs, paranormal and religious beliefs, vaccine hesitancy, and ideology in an underrepresented culture, Türkiye. We found that the correlated 4-factor model of 4-CTSQ is an appropriate measure to capture individual differences based on cognitive style. The results endorse the notion that cognitive style differences are characterized by distinct structures rather than being confined to two ends of a single continuum.

This article describes a study of transitivity of preference and of transparent dominance with 220 participants who judged each choice problem 4 times. It shows how a true-and-error model with 2 error terms per choice problem can be applied to replicated data, to ask if violations of dominance or of transitivity are ‘real’ rather than due to random response errors. These models allow one to estimate the incidence of systematic violations and of error rates. The new data showed about 3% violations of transitivity, corrected for error. This incidence might be statistically significant, but a skeptic might dismiss it as too small to build a theory upon. Tests of dominance found violations with overall rates from 4% to 18%. As in previous research, violations of transparent dominance appeared almost exclusively among people who systematically prefer ‘safe’ gambles (with low ranges of outcomes) over ‘riskier’ gambles with higher expected values when the ‘safe’ gamble was dominated by the higher-ranged gamble. For those participants and choice problems, rates of violation were 28%–45%, corrected for error. It was theorized that these violations of dominance may be due to a subgroup of risk-averse participants using a strategy in this experiment to find ‘safe’ alternatives, without comparing outcomes between gambles.