Skorinko et al. ran six experiments exploring gender and racial attitudes where participants believed they would interact with a partner and alignment between attitudes was assessed. Before each interaction, some participants were primed with a perspective-taking mindset and given information about their ostensible partner's views. The authors demonstrated that perspective-taking primes consistently influenced explicit attitudes, yet failed to impact implicit attitudes. Demand characteristics might have played a more influential role on the explicit level, which implicit measures are less susceptible to. These results call for further exploration into the nuanced mechanisms underlying implicit attitudes and the role of perspective-taking in shaping them.

Müller et al. investigated if non-political anxiety (electric shock) could affect political attitudes. Participants were randomly assigned to have their political attitudes assessed either under threat (shock) or when safe (no shock threat). Psychometric and physiological data confirmed successful state anxiety induction, but this induction did not alter implicit and explicit political attitudes. This research challenges prior assumptions about the relationship between anxiety and political ideology, highlighting the importance of contextual factors (political context of threat) in understanding the impact of threats on ideology. The findings also raise the question of whether trait anxiety, as opposed to state anxiety induced in this study, might play a more significant role in influencing implicit and explicit political attitudes, suggesting a direction for future studies to explore.

Marini and Casile explored how white participants embodied in either a Black or White Virtual Reality (VR) avatar, impacted their implicit racial bias. The Black avatar condition showed a reduction in participants' implicit pro-White/anti-Black attitudes compared to their baseline score, but only when they could see their virtual body from first-person and reflected perspectives. This suggests that the ability to see one's virtual body in a mirror/multiple perspectives may be crucial for the effectiveness of VR interventions in ameliorating implicit biases.

Lin et al. explored the relationship between parents' education level and the gender role characteristics of ideal mates among college-aged Chinese participants, considering the moderating role of urban-rural residence. Across two studies they showed that higher parental education levels were linked to female students preferring mates with high-femininity, low-masculinity traits. For male students, higher parental education levels were associated with explicit preferences for high-masculinity, low-femininity traits. The study concluded that parents with higher educational attainment might raise children who favor partners with non-traditional gender roles.

Qian et al. investigated whether differentiating among multiple (four Black instructors) or a single Black individual for 2 min is optimal for reducing implicit racial bias in 4-to-6-year-olds. Using a child-friendly IAT, they showed a reduction in implicit bias against Black people only in the differentiation condition, indicating that learning to distinguish among multiple other-race individuals is critical for reducing children's implicit racial bias.

Across six studies, Rubinstein et al., addressed the moderating impact of the perceived controllability of weight and religious social group membership on bias perception, finding no supportive evidence. However, individuating information was consistently effective at shifting implicit and explicit scores toward targets from existing social groups, but results were inconsistent for novel social groups. Overall, individuation was a robust and promising means of reducing implicit biases for existing social groups.

Hauser and Schwarz explored how neutral concepts, when frequently paired with valenced words, may foster implicit biases. They discovered that, unlike unpaired neutral primes, neutral primes paired with valenced words influenced participants' evaluations similarly to strongly valenced primes. This suggests that the implicit associations participants formed were driven not just by the direct meaning of words but by the consistent contexts in which these words appear.

Two studies found support for the causal embedding hypothesis, indicating that the language we encounter in daily life—specifically, the collocation patterns of words—can embed societal biases into our cognition. This means that linguistic biases may not only reflect societal biases but also play an active role in perpetuating them. They suggest that “anyone learning a culture's language may unwittingly learn that culture's implicit biases”.

Williams and Polito conducted two experiments to test if brief mindfulness meditation could reduce implicit age attitudes and sunk-cost decision-making (investment based on cumulative prior investment) bias, including increasing organizational citizenship behaviors (OCB; voluntary actions that benefit the organization and its members beyond formal job duties) at work. They showed that while mindfulness meditation significantly increased the intention to perform OCB, it did not reduce implicit age attitudes and sunk cost bias. Further research is needed to elucidate the effectiveness of meditation in bias reduction in professional environments.