A new study published in Scientific Reports provides a detailed model of how the human brain develops during the transition from the womb to early infancy. The findings indicate that distinct growth patterns for different brain tissues and sex-based differences in brain volume are established between mid-pregnancy and the first weeks of life. This research offers a continuous view of how the brain expands during a foundational period that was previously difficult to map.
The perinatal period involves rapid biological changes that establish the core architecture of the human brain. This phase includes the processes where cells proliferate, migrate to their correct locations, and begin forming complex connections. Scientists have often studied prenatal development and postnatal development separately because of the technical challenges involved in imaging fetuses compared to newborns. This separation has historically made it difficult to understand exactly how growth trajectories evolve as a fetus becomes an infant.
To bridge this gap, a research team led by the University of Cambridge aimed to create a unified model of early brain growth. Yumnah T. Khan, a PhD student at the Autism Research Centre at the University of Cambridge, led the investigation. The team sought to determine when specific tissues dominate growth and when sex differences in brain size first appear. By combining data from before and after birth, they hoped to capture the dynamic nature of brain structural changes.
The researchers utilized data from the Developing Human Connectome Project, which is a large-scale initiative designed to map brain connectivity. The final dataset included 798 magnetic resonance imaging scans collected from 699 unique individuals. These participants included 263 fetuses scanned while in the womb and 535 newborns.
The sample consisted of 380 males and 319 females. The scans covered a developmental window ranging from just over 21 weeks to nearly 45 weeks after conception. This allowed the team to track changes across the second and third trimesters of pregnancy and into the first month after birth.
The team used advanced statistical modeling to chart the volume of different brain tissues against the age of the individuals. They applied corrections to account for the natural variance that increases as infants grow older. The analysis focused on total brain volume as well as specific compartments like gray matter, white matter, and cerebrospinal fluid.
The analysis revealed that the total volume of the brain grows at an increasing rate leading up to birth. When the researchers accounted for the exact age at the time of the scan, they observed a slight slowing of this growth rate in the weeks immediately following birth. This suggests the most rapid expansion occurs just before and shortly after delivery.
Different types of brain tissue followed their own unique timelines. White matter, which forms the connections between brain cells, was the primary driver of growth during mid-pregnancy. However, its proportional contribution to the total brain size decreased over time. This suggests the brain prioritizes establishing core connectivity pathways early in gestation.
In contrast, gray matter, which contains the cell bodies of neurons and is involved in processing information, became the dominant driver of growth during late pregnancy and the postnatal period. This shift indicates a transition from laying down connections to the proliferation and maturation of processing centers. The rapid growth of gray matter likely supports the development of sensory and motor abilities needed for survival after birth.
The study also looked at deep brain structures known as subcortical regions. These areas, such as the amygdala and thalamus, showed an earlier peak in their growth rates compared to the outer layer of the brain, the cortex. The cortex is typically associated with higher-level cognitive functions.
The finding that subcortical structures mature faster aligns with the understanding that regions responsible for basic physiological and sensory functions develop before those involved in complex thought. The researchers observed that the cerebellum, a region critical for motor control, showed exponential growth throughout the studied period. This rapid expansion likely facilitates the early coordination required for an infant’s movements.
A major component of the analysis involved comparing brain development between males and females. The data showed that, on average, males experienced greater increases in brain volume as they aged compared to females. This difference was observable across the entire brain and within specific regions.
The researchers found that these sex differences were generally linear, meaning males consistently showed faster growth. This provides evidence that sex differences in brain structure are not solely a result of social or environmental influences after birth. Instead, biological factors present during pregnancy appear to initiate these divergence patterns.
While males exhibited faster overall growth, the shape of the growth trajectories was largely similar between the sexes. Both males and females followed the same general patterns of tissue expansion. However, there were specific exceptions in regional development.
For example, parts of the temporal lobe showed more pronounced gray matter increases in males. Additionally, the team identified a distinct growth pattern in the left anterior cingulate gyrus. In this region, males showed an S-shaped growth curve, whereas females showed a linear trajectory.
The study faces certain limitations regarding the available data. The scans for fetuses did not begin until after 21 weeks of gestation, leaving the first half of pregnancy unmapped in this analysis. Additionally, the number of scans available for younger fetuses was smaller than for older infants, which could impact the precision of the early growth models.
The researchers also noted technical differences between how fetal and neonatal scans were acquired. Although the same scanner was used, the settings had to be adjusted for the different environments of the womb and the nursery. This could potentially introduce variations in the measurements, though the team observed strong continuity in the data.
While the study documents when sex differences emerge, it does not confirm the biological mechanisms causing them. The authors suggest that prenatal hormones like testosterone likely play a role. Male fetuses are exposed to a surge of testosterone between 14 and 18 weeks of gestation.
The timing of the observed structural differences, appearing after 18 weeks, corresponds with the aftermath of this hormonal surge. Future research will need to directly investigate the link between hormone levels and these structural changes to confirm causality. The researchers emphasize that understanding these typical growth trajectories provides a baseline for identifying atypical development.
This baseline could eventually help explain why certain neurodevelopmental conditions are more common in one sex than the other. For instance, autism is diagnosed more frequently in males. Understanding if and how early brain overgrowth relates to these conditions remains a priority for the field.
The team calls for further longitudinal studies to validate these findings over longer periods. Following the same individuals from pregnancy through childhood would provide even stronger evidence for these developmental patterns. The current study represents a significant step toward a complete map of early human brain development.
A new analysis of gene expression in blood samples suggests that specific biological signs of Parkinson’s disease are detectable years before physical symptoms appear. These molecular signatures, related to how cells repair DNA and handle stress, seem to fade once the disease is fully established. The findings were published in npj Parkinson’s Disease.
Parkinson’s disease is traditionally diagnosed only after significant brain damage has occurred, typically manifested by tremors, stiffness, and slowness of movement. Scientists have long sought ways to identify the condition during the “prodromal” phase. This phase represents a period when internal biological changes are happening, but the classic motor symptoms have not yet surfaced. Identifying the disease at this stage is a major goal for medical science because it offers a potential window for early intervention.
Danish Anwer, a doctoral student at the Department of Life Sciences at Chalmers University of Technology in Sweden, led a team to investigate whether these early internal changes could be tracked in the blood. The research team operated on the hypothesis that the body’s genetic instructions for repairing DNA might be overactive or dysregulated early in the disease process.
Dopamine-producing neurons in the brain are high-energy cells that naturally produce toxic byproducts during their activity. These byproducts can damage DNA, requiring a robust repair system to keep the cells healthy.
The researchers theorized that in the early stages of Parkinson’s, these repair systems might be working overtime to save the dying cells. If this activity could be detected in the blood, it would serve as an early warning system. To test this, they needed to look at how these biological processes change over time rather than just taking a single snapshot.
The research team utilized data from the Parkinson’s Progression Markers Initiative, a large-scale observational study that tracks the evolution of the disease. They analyzed blood samples collected over a period of up to three years. The study included 188 healthy individuals to serve as a control group.
In addition to the healthy controls, the study analyzed 393 patients who had already been diagnosed with established Parkinson’s disease. Crucially, the researchers also included 58 individuals in the prodromal phase. These are people who do not yet have the motor symptoms of Parkinson’s but exhibit early warning signs such as REM sleep behavior disorder or loss of smell.
The researchers used a technique called RNA sequencing to look at the activity levels of thousands of genes in these blood samples. While DNA is the instruction manual, RNA is the message that tells the cell what to do at any given moment. By sequencing the RNA, the team could see which genes were being turned on or off.
They specifically examined genes responsible for three key biological pathways. The first was mitochondrial DNA repair, which maintains the energy generators of the cell. The second was nuclear DNA repair, which protects the main genetic code. The third was the integrated stress response, a safety mechanism cells use to handle dangerous conditions.
To analyze this vast amount of data, the team employed machine learning algorithms known as logistic regression classifiers. These computer models were trained to distinguish between the different groups based on their gene expression profiles. The researchers assessed how accurately these models could identify a person as healthy, prodromal, or having established Parkinson’s based solely on their blood data.
The investigation revealed that gene activity related to DNA repair and stress responses could accurately distinguish prodromal individuals from healthy controls. The models achieved high accuracy in identifying those in the early, pre-symptomatic stages. The accuracy of these predictions tended to improve as the participants moved closer to the typical time of diagnosis.
In contrast, these same gene patterns could not effectively separate patients with established Parkinson’s disease from healthy people. This suggests that the molecular signals are strong and distinct during the early development of the disease but quiet down later. Once the disease is clinically apparent, the gene expression in the blood appears to return to a state similar to that of healthy individuals.
The researchers observed that gene expression in the prodromal group was highly variable at the beginning of the study. Over the course of two to three years, this variability decreased significantly. This pattern indicates that the body initially mounts a chaotic or intense effort to repair cellular damage. As the disease progresses, this protective response appears to burn out or fail.
This concept was further supported by the observation of non-linear patterns in gene activity. About half of the DNA repair genes did not simply increase or decrease in a straight line. Instead, they followed complex trajectories, rising and then falling, or vice versa. This suggests a dynamic and transient biological struggle occurring before the onset of motor symptoms.
The study highlighted specific genes that were particularly predictive of the prodromal state. These included ERCC6 and NEIL2, both of which are involved in fixing damage to DNA. ERCC6 is known to be important for repairing active genes and is linked to conditions involving premature aging. NEIL2 helps repair damage caused by oxidative stress, which is a known factor in the death of dopamine neurons.
Another notable gene identified was NTHL1. This gene showed high importance as a predictor early in the prodromal phase. However, its relevance declined sharply as time passed. This decline supports the theory that specific repair mechanisms are recruited early on but eventually become overwhelmed or inactivated as the neurodegeneration advances.
The team also compared these specific stress and repair genes against broader sets of genes usually associated with Parkinson’s disease. They found that the repair and stress response genes were superior at identifying the prodromal phase. This indicates that general Parkinson’s risk genes might be less useful for tracking the active disease process in its earliest stages compared to these specific repair pathways.
The inability of the models to distinguish established Parkinson’s from controls is a significant finding. It implies that by the time a patient sees a doctor for tremors, the systemic battle in the blood has largely subsided. This highlights a limited temporal window where blood tests based on these markers would be effective.
There are limitations to this research that should be considered when interpreting the results. Blood samples serve as a proxy and do not always perfectly reflect what is happening inside the brain. It is possible that the signals detected in the blood are distinct from the specific degeneration occurring in central nervous system cells. The changes in the blood might reflect a systemic response to the disease rather than the direct brain pathology.
Additionally, the sample size for the prodromal group was relatively small compared to the other groups. While the statistical methods used were robust, larger studies will be necessary to confirm these patterns. The researchers also noted that external factors like medication could influence gene expression in established patients, potentially masking some signals.
The researchers did not perform functional tests to see if the changes in RNA levels resulted in changes in actual protein levels or cellular function. Gene expression is only the first step in protein production. Future studies will need to bridge the gap between these genetic signals and the actual cellular machinery.
Despite these limitations, the study provides evidence that the prodromal phase of Parkinson’s is biologically distinct from the established phase. It suggests that the body fights the disease aggressively in the beginning. This insight could help in the design of clinical trials by allowing researchers to select patients who are in this active, early phase.
The research team aims to understand exactly how these early repair mechanisms work and why they eventually fail. Developing these findings into a practical blood test for clinical use will require further testing and regulatory approval. The scientists estimate that such a test could potentially begin trials in healthcare settings within five years.
New research suggests that a college student’s level of narcissism plays a role in how they perceive and participate in flirtatious interactions with their professors. The findings indicate that students with high levels of grandiose narcissism are more likely to report flirting with faculty and believe faculty are flirting back, whereas those with vulnerable narcissism tend to perceive such behavior as common among their peers but not within their own interactions. The study was published in The Journal of Social Psychology.
The dynamics of student-professor relationships have long been a subject of concern within higher education. While most interactions remain professional, sexual or romantic engagements do occur and can lead to serious consequences. These include lawsuits, conflicts of interest, and the erosion of a safe learning environment.
Despite the gravity of these issues, there has been very little empirical research into which individual personality traits might predict the initiation of such behaviors. Previous research from the early 1980s suggested that a significant portion of students had flirted with professors, but modern data on the psychological drivers behind these actions has been sparse.
“While researchers are often interested in how narcissism influences behavior within academia, previous research has focused on academic success (e.g., GPA) and/or academic misconduct (e.g., cheating),” explained study author Braden T. Hall, a PhD student at the University of Alabama.
“However, flirting between students and professors is a real-world problem with serious consequences (e.g., damage to reputation, severe power imbalances, damage to academic integrity, lawsuits, etc.), and no research has examined the types of students that may be more likely to engage in such behavior, perceive such behavior from their professors, or perceive such behavior as prevalent on their campus and/or less morally inappropriate.”
Narcissism is generally understood as a personality trait characterized by a sense of entitlement and self-importance. However, psychologists recognize two distinct forms: grandiose and vulnerable.
Grandiose narcissism is associated with boldness, charm, and a desire for admiration. Vulnerable narcissism involves similar entitlement but is coupled with insecurity, anxiety, and a sense of victimization. The research team proposed that these two types of narcissism would manifest differently regarding academic flirting.
The researchers hypothesized that grandiose individuals would be bold enough to flirt personally, while both types would view the behavior as more acceptable and prevalent among others. To test their hypotheses, the researchers recruited 233 undergraduate psychology students from the University of Alabama.
The sample was predominantly female and white, with an average age of 19. Participants began by completing the Five-Factor Narcissism Inventory – Short Form, a standardized measure designed to assess levels of both grandiose and vulnerable narcissism. This allowed the team to score each participant on the specific dimensions of the personality trait.
The core of the study involved a detailed assessment of flirting behaviors. To ensure the behaviors listed were relevant, the researchers first conducted a pilot study to identify actions that students and faculty agreed constituted flirting. This resulted in a list of 12 specific behaviors for classroom settings, such as complimenting appearance, and 12 for office settings, such as sitting on a desk. Importantly, these behaviors were designed to be mild to moderate in nature rather than explicit sexual harassment or coercion.
Participants reviewed these behaviors and provided frequency estimates across several different scenarios. They rated how often they engaged in these behaviors toward professors and how often professors engaged in them toward the students. They also provided estimates for how often they believed their peers engaged in these behaviors with professors. Finally, the students rated the moral appropriateness of the behaviors. The researchers used statistical models to analyze how narcissism scores predicted these frequency estimates and moral judgments.
The results provided evidence that narcissism influences how students view academic boundaries. Students with higher levels of grandiose narcissism reported engaging in flirting behavior with professors more frequently. They also reported that professors flirted with them more often.
This pattern was consistent regardless of whether the interaction took place in a classroom or an office. This finding aligns with the profile of grandiose narcissists as individuals who seek attention, lack fear of social rejection, and may view themselves as exceptionally attractive or desirable to authority figures.
The findings for vulnerable narcissism were distinct. Students scoring high in vulnerable narcissism did not report higher frequencies of flirting with professors themselves. This is likely due to the social anxiety and fear of rejection that characterizes this form of narcissism. Although they may desire special treatment, the risk of awkwardness or dismissal likely inhibits them from acting on those desires.
However, vulnerable narcissism did predict how students viewed the behavior of others. High levels of vulnerable narcissism were associated with the belief that peers were frequently flirting with professors and that professors were flirting with peers. This suggests a cynical worldview where these students believe others are getting ahead through manipulative or immoral means, even if they are not doing so themselves.
When it comes to moral judgment, both forms of narcissism showed similar patterns. Higher levels of both grandiose and vulnerable narcissism were associated with viewing student-professor flirting as less inappropriate.
While the average student in the study viewed these behaviors as generally inappropriate, narcissistic students were more tolerant of them. This aligns with previous research suggesting that narcissism is linked to “moral disengagement,” or the tendency to excuse unethical behavior when it serves one’s interests or matches one’s worldview.
“Most of the effects of narcissism we found were medium-to-large, so these effects seem robust, and the effects of grandiose narcissism were consistent across contexts (e.g., classroom and offices), suggesting that these effects are due to trait-level differences rather than situations,” Hall told PsyPost.
The study also revealed general trends regarding the context of these interactions. Participants tended to view flirting as less inappropriate when it occurred in a classroom compared to a private office. The researchers suggest this might be because classroom interactions are public and may be interpreted as trying to be entertaining or engaging, whereas private office interactions imply a higher level of intimacy and potential for misconduct.
“Flirting between students and professors, while oftentimes seemingly benign, can be misinterpreted and have serious consequences in academic settings,” Hall explained. “The present study offers novel insight into the types of students (grandiose and vulnerable narcissistic students) who are more likely to see this behavior as less morally troubling and believe that flirting between students and professors is more typical. Additionally, we draw an important distinction wherein only grandiose narcissistic students are more likely to see flirting as typical of themselves.”
But it is important to contextualize these findings within the broader scope of the data. The average frequency estimates for flirting were low across the board. This means that while narcissistic students reported more flirting than their less narcissistic counterparts, the absolute reported frequency was still relatively rare.
Most students do not flirt with professors, and most view it as wrong. The study does not suggest that universities are overrun with flirtatious exchanges, but rather that when they do occur, specific personality traits are likely involved.
Even more narcissistic students “did not rate flirting between students and professors as appropriate, just less inappropriate,” Hall noted.
As with all research, there are also some limitations to consider. The research relied entirely on self-reported data. It is possible that grandiose narcissistic students merely believe they are flirting or being flirted with due to their inflated ego, rather than accurately reporting reality. The study was also cross-sectional, meaning it captured a snapshot in time and cannot definitively prove that narcissism causes the behavior, only that they are related.
Additionally, the sample was drawn from a large state university in the southeastern United States. “It would be interesting to see if these effects replicate at smaller universities where students and professors may have closer one-on-one relationships, which may lend itself to stronger effects,” Hall said.
A new study suggests that artificial intelligence systems approach strategic decision-making with a higher degree of mathematical optimization than human players, often outperforming humans in games requiring iterative reasoning. While these large language models demonstrate an ability to adapt to complex rules and specific competitive scenarios, they differ fundamentally from human cognition by failing to identify certain logical shortcuts known as dominant strategies. The findings appear in the Journal of Economic Behavior and Organization.
Large language models are advanced artificial intelligence systems designed to process and generate text based on vast datasets. These models are increasingly integrated into economic workflows, ranging from market analysis to automated negotiation agents. As these tools become more prevalent in settings that involve social interaction and competition, it becomes necessary to understand how their decision-making processes compare to human behavior.
Previous psychological and economic research indicates that humans often rely on bounded rationality, meaning their strategic thinking is limited by cognitive capacity and time. Iuliia Alekseenko, Dmitry Dagaev, Sofiia Paklina, and Petr Parshakov conducted this study to determine if artificial intelligence mirrors these human limitations or operates with a distinct form of logic. The authors are affiliated with HSE University, the University of Lausanne, and the New Economic School.
“This study was motivated by a growing debate about whether large language models can meaningfully serve as substitutes for human decision-makers in economic and behavioral research. While recent work has shown that LLMs can replicate outcomes in some classic experiments, it remains unclear how they reason strategically and whether their behavior truly resembles human bounded rationality,” the researchers told PsyPost.
“We focused on the beauty contest game because it is one of the most extensively studied tools for measuring strategic thinking and iterative reasoning in humans, with decades of experimental evidence across different populations and settings. This made it an ideal benchmark for a direct comparison between human behavior and AI-generated decisions.”
“More broadly, we were motivated by a real-world concern: AI systems are increasingly used in strategic environments such as markets, forecasting, and negotiation. Understanding whether AI models reason like humans, better than humans, or simply differently is crucial for predicting how they may influence outcomes when interacting with people.”
The researchers utilized a classic game theory experiment known as the “beauty contest” or “Guess the Number” game. In this game, participants simultaneously choose an integer between 0 and 100. The winner is the player whose chosen number is closest to a specific fraction of the average of all chosen numbers.
A common version sets the target at two-thirds of the average. If all players chose numbers randomly, the average would be 50, and the target would be 33. A sophisticated player anticipates this and chooses 33. If all players are equally sophisticated, they will all choose 33, making the new target 22. This reasoning process repeats iteratively until it reaches 0, which is the theoretical Nash equilibrium.
To test the capabilities of artificial intelligence, the authors employed five prominent large language models: GPT-4o, GPT-4o-Mini, Gemini-2.5-flash, Claude-Sonnet-4, and Llama-4-Maverick. The researchers replicated sixteen distinct scenarios from classic behavioral economics papers. These scenarios varied the number of players, the target fraction, and the aggregation method used to determine the winner.
The study gathered 50 responses from each model for every scenario to ensure statistical reliability. The temperature parameter for the models was fixed at 1.0 to allow for variability similar to a diverse group of human participants.
The study first replicated an experiment originally conducted by Rosemarie Nagel in 1995. The artificial agents played a version of the game where the target was either one-half or two-thirds of the average. In the scenario where the target was one-half, human participants typically chose numbers averaging around 27.
The artificial intelligence models consistently chose lower numbers. For example, the Llama model averaged a guess of 2.00, while Claude Sonnet averaged 12.72. This pattern persisted in the two-thirds variation. While humans averaged 36.73, the models provided mean guesses ranging from 2.80 to 22.24. This suggests that the models engaged in more steps of iterative reasoning than the average human participant.
The researchers also replicated a study by Duffy and Nagel from 1997 to see how the models handled different winning criteria. In this set of experiments, the winner was determined by being closest to one-half of the median, mean, or maximum of the chosen numbers. Human players tend to choose higher numbers when the target is based on the maximum.
The large language models successfully replicated this comparative static. When the target function changed to the maximum, models like Claude Sonnet and GPT-4o shifted their guesses upward significantly. This indicates that the models are capable of recognizing how changes in the rules should theoretically impact the optimal strategy.
A separate set of experiments focused on two-player games, initially studied by Grosskopf and Nagel in 2008. In a two-player game where the target is two-thirds of the average, choosing 0 is a weakly dominant strategy. This means that choosing 0 is never worse than any other option and is often better.
Despite this mathematical certainty, the models failed to identify the dominant strategy explicitly. The researchers analyzed the reasoning text generated by the models and found no instances where a model correctly explained the concept of a dominant strategy in this context. While the models played low numbers, they arrived at their decisions through probabilistic reasoning rather than by solving the game logically.
“Two things stood out,” the researchers said. “First, we were surprised by how consistently AI models behaved more strategically than humans across very different experimental settings. Second, and more unexpectedly, even the most advanced models failed to explicitly identify a simple dominant strategy in a two-player game, revealing an important gap between sophisticated-looking reasoning and basic game-theoretic logic.”
“Across many settings, AI models behaved much more strategically than humans, often choosing values far closer to the theoretical benchmark, which would meaningfully alter outcomes in real strategic interactions. At the same time, these effects highlight differences rather than superiority, since AI also shows clear limitations in recognizing simple dominant strategies.”
The researchers further investigated whether models could simulate specific human traits, replicating work by Brañas-Garza and colleagues. The prompts were adjusted to describe the artificial agent as having either high or low cognitive reflection scores. When instructed to act as an agent with high cognitive reflection, the models chose lower numbers. When instructed to act as an agent with low cognitive reflection, they chose higher numbers.
This alignment matches the behavioral patterns observed in actual human subjects. The models demonstrated a similar ability to simulate emotional states. When prompted to experience anger, the models chose higher numbers, mirroring findings from Castagnetti and colleagues that showed anger inhibits deep strategic reasoning in humans.
The researchers also examined the effect of model size on performance using the Llama family of models. They tested versions of the model ranging from 1 billion to 405 billion parameters. A clear correlation emerged between model size and strategic behavior.
The smaller models produced guesses that deviated substantially from the Nash equilibrium, often matching or exceeding human averages. The largest models produced results much closer to zero. This implies that as artificial intelligence systems scale in complexity, their behavior in strategic settings tends to converge toward the theoretical mathematical optimum rather than typical human behavior.
“A key takeaway is that modern AI systems can reason strategically and adapt to different situations, but they do not think in the same way humans do,” the researchers told PsyPost. “In our experiments, AI models consistently behaved in a more strategic and calculation-driven manner than people, even compared to well-educated or expert human participants.”
“At the same time, the study shows that AI reasoning is not simply a more advanced version of human reasoning. Despite their sophistication, the models failed to identify a basic dominant strategy in a simple two-player game, highlighting important limitations and blind spots.”
“For the average reader, this means that AI decisions should not be interpreted as direct predictions of human behavior. When AI systems are used in settings that involve judgment, competition, or social interaction, they may push outcomes in directions that differ from what we would expect if only humans were involved.”
There are some limitations to the study’s findings. The artificial agents were not playing for real financial incentives, which is a standard component of behavioral economics experiments with humans. The absence of a tangible reward could influence the depth of reasoning the models employ. Additionally, the study relied on specific phrasing in the prompts to simulate the experimental conditions. While robustness checks with paraphrased prompts showed consistent results, the models exhibited some sensitivity to how the task was framed.
“A common misinterpretation would be to conclude that AI thinks like humans or can be used as a direct proxy for human decision-making,” the researchers noted. “Our results show that while AI can perform well in strategic tasks, its reasoning patterns differ in important ways, and these differences can meaningfully affect outcomes. The key caveat is that strong performance in a task does not necessarily imply human-like cognition.”
“Our next step is to extend this approach to a wider set of strategic games that capture different cognitive demands, such as coordination, cooperation, and dominance reasoning. Ultimately, our goal is to build a systematic benchmark that compares human and AI behavior across multiple economic and psychological games, allowing researchers to better understand where AI aligns with human reasoning and where it diverges.”
A new neuroimaging study suggests that adolescents with borderline personality disorder exhibit distinct patterns of brain activity when reflecting on their own identity. The findings indicate that these young patients show reduced activation in the dorsolateral prefrontal cortex, a region associated with cognitive control, compared to healthy peers. This research was published in Translational Psychiatry.
Borderline personality disorder is a serious mental health condition. It is characterized by pervasive instability in moods, interpersonal relationships, self-image, and behavior. A central feature of this disorder is a disturbed sense of identity. Individuals often experience shifting goals, values, and vocational aspirations. This instability can manifest early in the course of the disorder.
Many previous studies have investigated the biological roots of the condition. Most of research has focused on emotional dysregulation rather than identity disturbance. Existing functional imaging studies have typically involved adult patients. These adult participants often have a history of medication use or co-occurring psychiatric conditions. These factors can make it difficult to determine which brain abnormalities are specific to borderline personality disorder itself.
To address this gap, a research team led by Pilar Salgado-Pineda from the FIDMAG Germanes Hospitalàries Research Foundation in Barcelona designed a study focusing on adolescents. They specifically sought participants who were in the early stages of the disorder. The team aimed to identify brain regions involved in the identity disturbance seen in the disorder. They focused on a developmental period that is critical for the formation of social cognition and self-concept.
The researchers recruited 27 female adolescents diagnosed with borderline personality disorder. These participants were between the ages of 12 and 18. Crucially, none of the patients had ever received pharmacological treatment for their condition. They were also screened to ensure they did not have any other comorbid psychiatric disorders.
For the control group, the researchers recruited 28 healthy female adolescents. These controls were matched to the patients in terms of age and estimated intelligence quotient. The strict selection criteria aimed to minimize confounding factors such as drug treatment and long-term illness effects.
The participants underwent functional magnetic resonance imaging. This technology measures brain activity by detecting changes associated with blood flow. While inside the scanner, the participants performed a task designed to engage self-reflection and reflection on others.
The task involved viewing a series of statements. Participants were asked to evaluate whether these statements were true or false. The statements belonged to one of three categories. The first category was the “self” condition, consisting of sentences about the participant. The second was the “other” condition, which involved sentences about an acquaintance the participant knew but was not emotionally close to.
The third category was a “facts” condition. This served as a control task and included general knowledge statements. The researchers also included a low-level baseline period where participants simply looked at a fixation cross on the screen. This design allowed the researchers to isolate brain activity specific to thinking about oneself and thinking about others.
The researchers analyzed the brain imaging data by comparing activation patterns between the different conditions. They specifically looked at the contrast between self-reflection and fact-processing. They also examined the contrast between other-reflection and fact-processing.
The analysis revealed differences in the group with borderline personality disorder during the self-reflection task. When comparing self-reflection to fact-processing, the healthy controls showed activation in several specific brain areas. These included the medial frontal cortex and the dorsolateral prefrontal cortex.
In contrast, the patients with borderline personality disorder showed reduced activation in the right dorsolateral prefrontal cortex. The patients also exhibited reduced activation in the left parietal cortex, the calcarine cortex, and the right precuneus.
The researchers conducted further analyses to understand the direction of these changes. They examined the activity levels in these regions relative to the fixation baseline. This revealed that while healthy controls activated the right dorsolateral prefrontal cortex during self-reflection, the patient group actually showed deactivation in this area.
The dorsolateral prefrontal cortex is widely recognized for its role in executive functions. It is heavily involved in top-down cognitive control. The authors suggest that the reduced activation in this region may reflect a diminished capacity for cognitive control over the process of self-reflection.
The study also examined brain activity during the other-reflection task. The results showed a different pattern of abnormality. When comparing other-reflection to fact-processing, the patient group appeared to show reduced activation in the medial frontal cortex. This region is part of the default mode network.
However, a detailed inspection of the data offered a nuanced explanation. The difference was not due to how the patients processed information about others. Instead, it was driven by a difference in the fact-processing condition. The healthy controls showed strong deactivation of the medial frontal cortex during the fact task. The patients failed to deactivate this region to the same extent.
The researchers interpret this specific finding as a failure of deactivation rather than a deficit in social cognition. This suggests that the brain mechanisms for thinking about others may be relatively preserved in these adolescents. The abnormality lay in the inability to suppress certain brain networks during a factual cognitive task.
The study notably found no differences between the groups in the temporoparietal junction. This brain region is known to be involved in understanding the beliefs of others. The lack of difference here implies that some aspects of social cognition might function normally in adolescents with the disorder.
There are limitations to this study that contextualize the findings. The sample included only female participants. Borderline personality disorder is diagnosed more frequently in females, but it does affect males. The findings may not extend to male adolescents with the condition.
The sample size was relatively small, with fewer than 30 participants in each group. Neuroimaging studies often require larger samples to detect subtle effects reliably. The strict exclusion criteria also limit generalizability. Most people with borderline personality disorder have other mental health conditions. Studying a “pure” sample helps isolate biological mechanisms but may not reflect the typical clinical population.
The study also relied on a specific experimental task to measure self-reflection. While this task is established in the field, it serves as an indirect measure of identity disturbance. The researchers did not include a behavioral measure of identity problems to correlate with the brain data.
Future research is needed to replicate these findings in larger and more diverse groups. Longitudinal studies could be particularly informative. Tracking adolescents over time would help clarify whether these brain activity patterns predict the worsening or improvement of symptoms as they enter adulthood.
A new study published in the journal Social Cognitive and Affective Neuroscience provides evidence that the human brain processes romantic partners differently than close friends, specifically within the reward system. The research suggests that while the brain creates a unique neural signature for a partner early in a relationship, this distinction tends to fade as the bond matures. These findings offer insight into how the biological drivers of romantic love may evolve from passion to companionship over time.
Relationships involve complex psychological states that differentiate a committed partner from a platonic friend. Scientists have sought to map these differences in the brain to understand the biological foundations of human bonding. Much of this research focuses on the nucleus accumbens. This small region deep within the brain, which relies heavily on the neurotransmitter dopamine, plays a central role in processing rewards and motivation.
Evidence from animal studies indicates that the nucleus accumbens is essential for forming pair bonds. Research on monogamous prairie voles shows that neurochemical signaling in this area drives the preference for a specific partner. The brain appears to undergo plastic changes that reinforce the bond.
Human studies have attempted to replicate these findings by comparing brain activity in response to partners versus friends. However, the results have been inconsistent. Some experiments observed higher activity in the nucleus accumbens for partners, while others found no significant difference. This inconsistency might stem from the fact that opposite-sex friends can sometimes be viewed as potential romantic alternatives.
“Romantic relationships are typically characterized by exclusivity, strong commitment, and passionate love, which distinguish them from friendships,” said study author Kenji Fujisaki
of the Department of Psychology at Kyoto University.
“We aimed to identify the neural mechanisms that distinguish romantic partners from friends. In addition, as romantic relationships develop, most people experience psychological fluctuations over time, raising the question of how neural processing of a partner may change as a relationship matures. Finally, given prior theory and evidence that opposite-sex friends can sometimes be potential or alternative partners, we were interested in whether the brain represents an opposite-sex friend more similarly to a romantic partner or to a same-sex friend.”
The study involved 47 heterosexual male participants. All participants were between the ages of 20 and 29 and were currently in a romantic relationship. The average length of these relationships was approximately 18 months. The researchers excluded individuals who were married or had children to control for the effects of long-term domestic partnership or parenthood.
To ensure the study captured genuine social bonds, the participants selected their own close friends to be part of the stimuli. They chose a close female friend and a close male friend. These friends, along with the romantic partners, provided short video clips for the experiment.
The researchers used functional magnetic resonance imaging to monitor brain activity while participants engaged in a specific activity called the social incentive delay task. This task is designed to measure the anticipation of a social reward. Participants saw a cue on a screen indicating which person would appear.
After a short delay, a target appeared on the screen for a fraction of a second. Participants had to press a button as quickly as possible. If they were successful, they saw a video clip of their partner, female friend, or male friend smiling and making a positive gesture. These gestures included waving, clapping, or making a “V-sign.”
If the participants were too slow, they saw a neutral expression instead. This design allowed the researchers to isolate the brain activity associated with anticipating social approval from specific people. The team analyzed the imaging data using a technique known as multivoxel pattern analysis.
Standard analysis looks at whether a brain region is “on” or “off.” In contrast, multivoxel pattern analysis examines the specific pattern of activity across many small segments of brain tissue. This allows researchers to see if the “neural fingerprint” for one person differs from another, even if the overall activity level is the same.
The behavioral results showed that the men were highly motivated by their partners. Participants reacted faster when anticipating a video of their partner compared to either friend. They also rated the videos of their partners as more likeable than those of their friends.
The brain imaging results revealed that the nucleus accumbens encodes the romantic partner in a distinct manner. The computer algorithms used in the analysis successfully differentiated the brain activity patterns associated with the partner from those associated with the female friend. This discrimination was possible across the nucleus accumbens and other related brain structures.
The researchers then assessed the similarity of these neural patterns. They found that in the nucleus accumbens, the representation of the female friend was more similar to the male friend than to the partner. This suggests that the brain categorizes the partner as a unique social entity, distinct from the general category of friendship.
A key finding emerged regarding the duration of the romantic relationships. The researchers analyzed whether the distinctiveness of the partner’s neural signature was related to how long the couple had been together. They observed a negative correlation between relationship length and neural specificity.
Participants who had been in their relationships for a longer time showed less distinct neural differences between their partner and their female friend. In the nucleus accumbens, the unique pattern that separated the partner from the friend appeared to diminish as the relationship length increased. This trend remained statistically significant even after the researchers controlled for self-reported levels of intimacy, passion, and commitment.
These results align with psychological theories describing the trajectory of love. Early stages of romance are often characterized by “passionate love,” which involves intense longing and motivation. This stage likely requires highly specific activity within the brain’s reward system to facilitate bond formation.
As a relationship stabilizes, it often transitions into “companionate love.” This form of love is characterized by deep attachment and friendship. The findings suggest that as this transition occurs, the biological processing of the partner in the reward system becomes less distinguishable from that of a close friend.
This reduction in neural distinctiveness does not imply a decline in the quality of the relationship. It may instead reflect a shift in how the relationship is biologically maintained. The intense, reward-driven signaling required to establish a bond may be less necessary for maintaining a stable, long-term union.
“Our results suggest that the way the brain represents a romantic partner is not fixed, but can evolve as a relationship develops,” Fujisaki told PsyPost. “Early in relationships, a reward-related brain region called the nucleus accumbens showed clearly differentiated activity patterns for a partner compared with an opposite-sex friend. In longer relationships, this neural distinction became less pronounced. This change may reflect a shift from the passionate love characteristic of early-stage relationships toward a more stable, companionate form of love that shares features with close friendship.”
As with all research, there are some limitations. The research relied on cross-sectional data. This means it compared different people at different relationship stages rather than following the same individuals over time. Longitudinal studies would be necessary to confirm that these changes occur within the same person.
The sample consisted entirely of heterosexual males. This decision was made to reduce biological variability in the sample. However, it limits the ability to generalize the findings to women or individuals with different sexual orientations. Future research needs to include more diverse samples to see if these neural patterns are universal.
The study focused primarily on the nucleus accumbens and the dorsal striatum. While these areas are central to reward, other brain regions are involved in social bonding. Areas responsible for emotional regulation or cognitive processing may take on a larger role in long-term relationships.
There is also the potential for misinterpretation regarding the “reduced specificity” finding. “A common misinterpretation would be to assume that reduced neural distinctiveness means that love or relationship quality is declining,” Fujisaki said. “Our findings do not support this conclusion, and the observed pattern should be understood as a group-level tendency that may vary across individuals.”
Future research could investigate identifying these complementary brain systems. It would be valuable to understand what neural mechanisms support enduring bonds once the specific reward processing in the nucleus accumbens becomes less pronounced. Additionally, examining how major life transitions like cohabitation or marriage affect these patterns could provide further insight.
“This study raises a new question: if partner-specific processing in the nucleus accumbens becomes less distinct over time, what neural mechanisms help sustain long-term relationships?” Fujisaki explained. “Moving forward, it would be worth identifying complementary brain systems that support enduring bonds.”
“In addition, further developing this work by examining neural processes underlying cognition and behavior characteristic of romantic relationships, while taking individual differences into account, may deepen our understanding of romantic bonding. Ultimately, this line of research could provide insights that help foster healthier and more satisfying romantic relationships.”
More than two years ago, Danish psychiatrist Søren Dinesen Østergaard published a provocative editorial suggesting that the rise of conversational artificial intelligence could have severe mental health consequences. He proposed that the persuasive, human-like nature of chatbots might push vulnerable individuals toward psychosis.
At the time, the idea seemed speculative. In the months that followed, however, clinicians and journalists began documenting real-world cases that mirrored his concerns. Patients were developing fixed, false beliefs after marathon sessions with digital companions. Now, the scientist who foresaw the psychiatric risks of AI has issued a new warning. This time, he is not focusing on mental illness, but on a potential degradation of human intelligence itself.
In a new letter to the editor published in Acta Psychiatrica Scandinavica, Østergaard argues that academia and the sciences are facing a crisis of “cognitive debt.” He posits that the outsourcing of writing and reasoning to generative AI is eroding the fundamental skills required for scientific discovery. The commentary builds upon a growing body of evidence suggesting that while AI can mimic human output, relying on it may physically alter the brain’s ability to think.
Østergaard’s latest writing is a response to a letter by Professor Soichiro Matsubara. Matsubara had previously highlighted that AI chatbots might harm the writing abilities of young doctors and damage the mentorship dynamic in medicine. Østergaard agrees with this assessment but takes the argument a step further. He contends that the danger extends beyond mere writing skills and strikes at the core of the scientific process: reasoning.
The psychiatrist acknowledges the utility of AI for surface-level tasks. He notes that using a tool to proofread a manuscript for grammar is largely harmless. However, he points out that technology companies are actively marketing “reasoning models” designed to solve complex problems and plan workflows. While this sounds efficient, Østergaard suggests it creates a paradox. He questions whether the next generation of scientists will possess the cognitive capacity to make breakthroughs if they never practice the struggle of reasoning themselves.
To illustrate this point, he cites the developers of AlphaFold, an AI program that predicts protein structures. This technology resulted in the 2024 Nobel Prize in Chemistry for researchers from Google DeepMind and the University of Washington.
Østergaard argues that it is not a given that these specific scientists would have achieved such heights if generative AI had been available to do their thinking for them during their formative years. He suggests that scientific reasoning is not an innate talent. It is a skill learned through the rigorous, often tedious practice of reading, thinking, and revising.
The concept of “cognitive debt” is central to this new warning. Østergaard draws attention to a preprint study by Kosmyna and colleagues, titled “Your brain on ChatGPT.” This research attempts to quantify the neurological cost of using AI assistance. The study involved participants writing essays under three conditions: using ChatGPT, using a search engine, or using only their own brains.
The findings of the Kosmyna study provide physical evidence for Østergaard’s concerns. Electroencephalography (EEG) monitoring revealed that participants in the ChatGPT group showed substantially lower brain activation in networks typically engaged during cognitive tasks. The brain was simply doing less work. More alerting was the finding that this “weaker neural connectivity” persisted even when these participants switched to writing essays without AI.
The study also found that those who used the chatbot had significant difficulties recalling the content of the essays they had just produced. The authors of the paper concluded that the results demonstrate a pressing matter of a likely decrease in learning skills. Østergaard describes these findings as deeply concerning. He suggests that if AI use indeed causes such cognitive debt, the educational system may be in a difficult position.
This aligns with other recent papers regarding “cognitive offloading.” A commentary by Umberto León Domínguez published in Neuropsychology explores the idea of AI as a “cognitive prosthesis.” Just as a physical prosthesis replaces a limb, AI replaces mental effort. While this can be efficient, Domínguez warns that it prevents the stimulation of higher-order executive functions. If students do not engage in the mental gymnastics required to solve problems, those cognitive muscles may atrophy.
Real-world examples are already surfacing. Østergaard references a report from the Danish Broadcasting Corporation about a high school student who used ChatGPT to complete approximately 150 assignments. The student was eventually expelled. While this is an extreme case, Østergaard notes that widespread outsourcing is becoming the norm from primary school through graduate programs. He fears this will reduce the chances of exceptional minds emerging in the future.
The loss of critical thinking skills is not just a future risk but a present reality. A study by Michael Gerlich published in the journal Societies found a strong negative correlation between frequent AI tool usage and critical thinking abilities. The research indicated that younger individuals were particularly susceptible. Those who frequently offloaded cognitive tasks to algorithms performed worse on assessments requiring independent analysis and evaluation.
There is also the issue of false confidence. A study published in Computers in Human Behavior by Daniela Fernandes and colleagues found that while AI helped users score higher on logic tests, it also distorted their self-assessment. Participants consistently overestimated their performance. The technology acted as a buffer, masking their own lack of understanding. This creates a scenario where individuals feel competent because the machine is competent, leading to a disconnect between perceived and actual ability.
This intellectual detachment mirrors the emotional detachment Østergaard identified in his earlier work on AI psychosis. In his previous editorial, he warned that the “sycophantic” nature of chatbots—their tendency to agree with and flatter the user—could reinforce delusions. A user experiencing paranoia might find a willing conspirator in a chatbot, which confirms their false beliefs to keep the conversation going.
The mechanism is similar in the context of cognitive debt. The AI provides an easy, pleasing answer that satisfies the immediate need of the user, whether that need is emotional validation or a completed homework assignment. in both cases, the human user surrenders their agency to the algorithm. They stop testing reality or their own logic against the world, preferring the smooth, frictionless output of the machine.
Østergaard connects this loss of human capability to the ultimate risks of artificial intelligence. He cites Geoffrey Hinton, a Nobel laureate in physics often called the “godfather of AI.” Hinton has expressed concerns that there is a significant probability that AI could threaten humanity’s existence within the next few decades. Østergaard argues that facing such existential threats requires humans who are cognitively adept.
If the population becomes “cognitively indebted,” reliant on machines for basic reasoning, the ability to maintain control over those same machines diminishes. The psychiatrist emphasizes that we need humans in the loop who are capable of independent, rigorous thought. A society that has outsourced its reasoning to the very systems it needs to regulate may find itself ill-equipped to handle the consequences.
The warning is clear. The convenience of generative AI comes with a hidden cost. It is not merely a matter of students cheating on essays or doctors losing their writing flair. The evidence suggests a fundamental change in how the brain processes information. By skipping the struggle of learning and reasoning, humans may be sacrificing the very cognitive traits that allow for scientific advancement and independent judgment.
Østergaard was correct when he flagged the potential for AI to distort reality for psychiatric patients. His new commentary suggests that the distortion of our intellectual potential may be a far more widespread and insidious problem. As AI tools become more integrated into daily life, the choice between cognitive effort and cognitive offloading becomes a defining challenge for the future of human intelligence.
A new study published in the Personality and Social Psychology Bulletin suggests that having a sense of power in a relationship promotes sexual assertiveness, while perceiving a partner as powerful fosters a willingness to accommodate their needs. The findings indicate that healthy sexual dynamics are not about one person holding dominance over another. Instead, the most satisfying interactions appear to occur when both individuals feel they have influence within the relationship.
Power dynamics are frequently viewed as potential sources of conflict or exploitation within intimate relationships. A common assumption is that if one partner holds power, they might satisfy their own desires while neglecting their partner. Alternatively, the partner with less power might feel forced to comply with unwanted activities.
“Power is commonly thought of as dangerous, particularly within sexual relationships,” said study author Nickola Overall, a professor at the University of Auckland and head of the REACH Lab. ”
“People who have high power in relationships might assert their own sexual need while neglecting their partner’s desires. But lacking power is also problematic. People who have low power in relationships might inhibit their desires and comply to undesired sexual activity. Despite these negative implications of having power and lacking power, how power relates to sexual assertiveness, neglect, and compliance is unclear.”
The researchers sought to clarify how a person’s own sense of power and their perception of their partner’s power distinctly shape sexual motivations and behaviors. They applied a theoretical framework that separates power into two distinct processes.
The first is “actor power,” or the individual’s own perceived ability to influence outcomes. The second is “perceived partner power,” or the individual’s belief in their partner’s ability to influence outcomes. The researchers proposed that one’s own power drives the decision to approach or inhibit sexual desires. Simultaneously, the perception of a partner’s power drives the decision to accommodate or neglect the partner’s needs.
“Most frameworks assume that one partner higher in power will be more assertive in pursuing their sexual needs in ways that neglect the other partner who will be pressured to comply,” Overall explained. “These frameworks assume that power is zero-sum in relationships – if one person has more power, then the other person has less power.”
“But relationships can involve both people having high power (mutually influencing each other), both having low power (lacking influence over each other), or one having more power than the other. And each person’s power can influence their behavior for potential good or ill.”
“All prior studies have only focused on one type of behavior, such as sexual assertiveness or sexual compliance, making assumptions about how these behaviors are linked, such as partners high in power asserting their needs risking the other person complying to undesired sexual activity. But, these distinct behaviors may be shaped by different processes and do not provide a full picture of people’s sexual relationships.”
“So we examined various outcomes relevant to different theories of power, including sexual assertiveness (e.g., comfort initiating sex), sexual compliance (e.g., agreeing to engage in undesired sexual activity), and sexual accommodation vs. neglect (e.g., being more vs. less willing to compromise and being more vs. less understanding when partners are not in the mood),” Overall said.
The research team conducted three separate studies. The first study involved 270 participants recruited from an online platform. These individuals were in committed, mixed-gender relationships and were currently childfree. The sample included 130 women and 140 men. Participants completed the Sense of Power Scale to rate their own ability to influence their partner. They responded to statements such as “I think I have a great deal of power.” They also completed a version of the scale assessing their partner’s power.
In this first study, participants also rated their comfort with initiating and refusing sex. They responded to direct statements like “I am comfortable initiating sex.” Additionally, they reported their history of compromising on sexual frequency or activities over the past six months.
The data showed that individuals who felt they had more power reported greater comfort in both initiating and refusing sexual intimacy. In contrast, those who perceived their partners as having more power expressed a higher willingness to compromise on sexual matters. The results suggested two separate pathways. One pathway leads to personal assertiveness. The other pathway leads to responsiveness to a partner.
The second study aimed to validate these initial observations with a more detailed methodology. The researchers recruited 152 couples, totaling 304 participants. This design allowed the team to analyze data from both partners in a relationship. The study included the same power measures as the first study but added the Hurlbert Index of Sexual Assertiveness. This index measures how openly participants express sexual needs. It includes items such as “I communicate my sexual desires to my partner.”
The second study also assessed sexual compliance. This construct refers to engaging in unwanted sexual activity. Participants rated items such as “I find myself having sex when I do not really want it.” Additionally, the researchers measured sexual communal strength. This is defined as the motivation to meet a partner’s needs. Participants answered questions regarding how far they would go to meet their partner’s sexual desires.
The findings from the second study reinforced the distinction between the two types of power. Participants with higher personal power scores reported higher levels of sexual assertiveness. Perhaps more importantly, those with lower personal power scores reported higher levels of sexual compliance. This suggests that engaging in unwanted sex is often driven by a lack of personal agency rather than the pressure of a powerful partner.
On the other hand, viewing a partner as powerful was linked to greater communal strength. This indicates that perceiving a partner as powerful motivates individuals to meet that partner’s needs rather than simply submit to them out of fear.
The third study expanded the scope further with a sample of 412 individuals recruited online. This iteration aimed to replicate the previous findings and introduce new measures. The researchers assessed “sexual acquiescence,” which captures participation in specific sexual acts without desire but without coercion.
They also measured reactions to sexual rejection. The team wanted to see if high power might lead to “sexual enticement,” or nagging a partner who has refused sex. They also measured “sexual understanding,” which involves accepting a partner’s lack of desire without negative feelings.
Consistent with the previous studies, high personal power predicted assertiveness. Low personal power predicted engaging in unwanted sex. Perceiving a partner as powerful predicted reacting to sexual rejection with understanding rather than persistence. The study found no evidence that high power leads to pressuring behaviors like enticement. This challenges the idea that powerful individuals inevitably use their influence to coerce partners.
Across all three studies, the researchers tested whether the effects differed between men and women. The analysis showed that the fundamental links between power and behavior were consistent regardless of gender. While men reported higher baseline levels of assertiveness and women reported higher compliance, the way power influenced these behaviors was the same for both groups. For both men and women, feeling powerful enabled them to say “no” when they wanted to. For both groups, seeing their partner as influential motivated them to be accommodating.
The researchers also examined “asymmetries,” or whether having more power than one’s partner caused specific issues. The results offered little evidence that power imbalances were the primary driver of behavior. The findings suggest that the combination of high actor power and high perceived partner power may yield the best outcomes. In this scenario, individuals feel free to express their own desires while simultaneously caring for their partner’s needs.
“Both people having power in relationships is important for people to enjoy a fulfilling sex life,” Overall told PsyPost. “When people lack power in their relationships—people feel unable to influence their partner—they are more likely to inhibit their sexual desires, such as being less comfortable in initiating sex or expressing their sexual needs and more likely to engage in sexual activity they do not desire. Sexual inhibition and compliance undermine people’s health and wellbeing, but also restrict the development of satisfying, connected relationships.”
“When partners lack power in relationships—people feel their partner is unable to influence them—they are more likely to neglect their partners’ needs, such as being less willing to compromise with their partner about when and how they have sex or being less understanding when their partner is not in the mood. Neglecting partners’ needs will harm both people in relationships because couples need to accommodate each other’s needs and desires to have fulfilling satisfying sex lives.”
“In short, healthy sexual relationships involve people being able to satisfy their own desires while accommodating their partner’s needs and desire. Hitting this sweet spot requires both partners having power in their relationship.”
These new findings align closely with recent research by Robert Körner and Astrid Schütz, which challenged the idea that power in relationships is a zero-sum game. In their studies published in The Journal of Sex Research and Social Psychological and Personality Science, Körner and Schütz established that relationship quality and sexual satisfaction hinge on an individual’s absolute sense of power rather than a perfect balance of power between partners.
The current study builds on this foundation by mapping these power dynamics to specific behavioral outcomes. While Körner and Schütz demonstrated that feeling powerful predicts positive sexual motivation, the new results explain how this functions: personal power drives the confidence to assert needs, whereas perceiving a partner as powerful drives the motivation to be generous and accommodating.
Both sets of research converge on the conclusion that high mutual power is preferable to power asymmetries or shared powerlessness. Körner and Schütz found that having a powerful partner does not diminish one’s own satisfaction, and similarly, the current study found no evidence that power imbalances are the primary driver of harmful behaviors like sexual compliance or neglect. Instead, both lines of inquiry suggest that the healthiest sexual dynamics occur when both partners feel a high sense of agency.
The new findings also offer a behavioral explanation for the profiles identified by Roxanne Bolduc and her colleagues in the Journal of Sex & Marital Therapy. Bolduc’s research indicated that individuals with egalitarian views and flexible preferences experience greater relationship satisfaction than those adhering to rigid or conflicted gender roles.
The current study supports this by demonstrating that the psychological mechanisms of power function similarly for men and women. By showing that high actor power promotes assertiveness and high partner power promotes accommodation regardless of gender, the findings illustrate why egalitarian dynamics, where both partners exercise influence, likely lead to the superior relationship outcomes observed in Bolduc’s “flexible” profile.
While the new findings provide insight into relationship dynamics, the study relies on self-reported data. Participants may not accurately report or be fully aware of their own behaviors. This is particularly true regarding sensitive topics like compliance or enticement. The cross-sectional nature of the data also prevents drawing definitive conclusions about cause and effect. It is possible that engaging in specific sexual behaviors influences a person’s sense of power, rather than the other way around.
Future research could benefit from longitudinal designs to track these dynamics over time. The samples consisted largely of people in established, committed relationships. Power dynamics might function differently in casual dating scenarios or relationships characterized by severe conflict. In contexts with less commitment, power imbalances might carry more risk of negative outcomes than observed in this study. Additionally, experimental studies could help clarify whether shifting a person’s sense of power directly causes changes in their sexual behaviors.
“Some perspectives warn that power can be dangerous by providing the opportunity to exploit low power others,” Overall added. “Our data show that in close relationships having power is likely to be more beneficial than harmful. People who felt they had power to influence their partner were more assertive in expressing their sexual needs and less compliant to unwanted sexual activity, but they were not less willing to compromise with their partners or less understanding when their partners were not in the mood. Similarly, people who perceived their partner had high power were more willing to compromise with their partner and less likely to neglect their partner’s needs, but they were not more likely to comply to unwanted sexual activity.”
“Many perspectives also suggest that power asymmetries are critical—one person having more power than the other risks greater neglect and compliance. But testing interactions between people’s own and their partners’ power did not provide any evidence for this. Instead, the few interactions that emerged suggested that jointly holding power solidified rather than reduced the positive effects of power – greater assertiveness in expressing sexual needs and accommodation of the partners’ sexual desires and lower compliance and partner neglect.”
“That said, our investigation examined power and sexual behavior within long-term intimate relationships in which both people care about and have some power over each other,” Overall continued. “In non- intimate contexts, like the workplace, one person holding power over another who has little or no counterpower could produce particularly harmful dynamics in which the person high in power can assert their needs while neglecting the other who may be more likely to comply. The risk of these harmful outcomes could also be greater in younger samples and dating couples that are not yet committed to one another, or in contexts where greater asymmetries between men and women restrict women’s power and sexual behavior.”
Distinct roles of power: The researchers identified two separate processes: “actor power” (a person’s own sense of influence), which drives sexual assertiveness and the confidence to refuse sex, and “perceived partner power” (a person’s view of their partner’s influence), which motivates a willingness to accommodate and compromise on the partner’s needs.
The “sweet spot” for satisfaction: Contrary to the idea that power in relationships is a zero-sum game where one person dominates the other, the findings suggest that the best sexual dynamics occur when both partners feel influential. This mutual power allows individuals to pursue their own desires while simultaneously caring for their partner’s needs.
Roots of sexual compliance: The study found that engaging in unwanted sexual activity (compliance) is primarily driven by a lack of personal agency (low actor power) rather than pressure from a powerful partner. Individuals who feel powerless are more likely to inhibit their own desires and agree to sex they do not want to avoid conflict.
Gender consistency: The link between power and sexual behavior was consistent for both men and women. Regardless of gender, feeling powerful facilitated boundary-setting and assertiveness, while perceiving a partner as influential fostered a motivation to be understanding and responsive to that partner.
Alignment with previous research: These findings reinforce other recent studies suggesting that relationship satisfaction depends on high absolute levels of power for both partners rather than just an equitable balance. The research supports the notion that egalitarian dynamics, where both parties exercise influence, produce better outcomes than rigid or conflicted gender roles.
A new study published in Translational Psychiatry has found that post-traumatic stress disorder is associated with accelerated biological aging in the brain. Researchers found that World Trade Center responders with PTSD had brains that appeared approximately three years older than their chronological age compared to responders without the disorder. This research suggests that the condition involves tangible structural changes to the brain that persist long after the initial trauma.
The health impacts of the September 11, 2001 attacks extend well beyond the immediate physical injuries sustained at Ground Zero. Many responders who assisted in the rescue and recovery efforts developed chronic psychological conditions. PTSD remains particularly prevalent in this population. Previous studies have linked the disorder to various markers of accelerated aging in the body, such as changes in immune function and inflammation.
The specific impact of the disorder on the biological aging of the brain itself has remained less clear. Determining how PTSD affects brain structure is necessary for understanding long-term health risks. Individuals with the condition face a higher statistical likelihood of developing age-related conditions like memory decline or dementia earlier in life. By identifying biological markers of brain aging, scientists hope to create better tools for early diagnosis and treatment.
“Nearly a quarter of World Trade Center responders continue to experience chronic PTSD more than two decades after 9/11, yet we still lack clear biological markers that capture its long-term impact on the brain,” said study author Azzurra Invernizzi of the Icahn School of Medicine at Mount Sinai.
“Previous MRI studies showed structural and functional brain differences in responders with PTSD, but these findings were often region-specific and difficult to translate into an overall picture of brain health. We wanted to address this gap by asking whether PTSD is associated with accelerated brain aging — a single, intuitive metric that reflects cumulative brain
changes and may help explain long-term cognitive and health risks in this population.”
The research team recruited 99 World Trade Center responders to participate in the study. This group included 47 individuals diagnosed with PTSD and 52 individuals with no history of the disorder. The participants were matched based on key demographics such as age, sex, and occupation to ensure a fair comparison. The average age of the participants was approximately 55 years.
Each participant underwent a high-resolution structural magnetic resonance imaging scan. The researchers then employed a specialized artificial intelligence tool called BrainAgeNeXt to analyze these scans. This tool uses a form of deep learning called a convolutional neural network. The model estimates a person’s “brain age” based on anatomical features captured in the MRI data.
The model was previously trained on over 11,000 MRI scans from healthy individuals to learn what a brain typically looks like at different stages of life. This training allows the software to bypass manual measurements and identify complex patterns across the entire brain volume. The team calculated a metric known as the Brain Age Difference for each responder.
This number represents the gap between the age predicted by the MRI scan and the person’s actual chronological age. A positive number indicates the brain appears older than expected. A negative number suggests it appears younger or consistent with healthy aging. The researchers used this metric to compare the two groups of responders.
“Brain age is a summary measure, not a diagnosis, but even modest shifts are meaningful because they reflect widespread changes across the brain rather than isolated regions,” Invernizzi explained. “Accelerated brain aging has been linked in other studies to cognitive decline and increased risk for age-related neurological conditions.”
The analysis revealed a significant distinction between the groups. Responders diagnosed with PTSD showed an average Brain Age Difference of approximately 3.07 years. In contrast, responders without the disorder showed an average difference of negative 0.43 years. This indicates that the brains of those with the condition showed structural signs associated with advanced age compared to their trauma-exposed peers.
“One striking aspect was how clearly PTSD status alone distinguished brain aging trajectories, even among individuals with shared exposures and similar demographic characteristics,” Invernizzi told PsyPost. “This suggests that PTSD itself may play a central role in shaping long-term brain outcomes, beyond general stress or aging effects.”
Further examination linked these higher brain age estimates to specific anatomical changes. The researchers observed associations between increased brain age and larger volumes of cerebrospinal fluid and ventricular spaces. These patterns typically signify a loss of brain tissue or atrophy. In the PTSD group specifically, a smaller thalamus was associated with an older-appearing brain. The thalamus is a region involved in sensory processing and fear regulation.
The study also assessed the duration of time responders spent working at the World Trade Center site. The data indicated that the length of exposure moderated the relationship between the disorder and brain age. Responders with PTSD who spent more time working at the disaster site tended to show greater increases in estimated brain age.
This interaction suggests that the combination of the psychological condition and prolonged exposure to the environmental stressors of the site may compound the effects on brain structure. Responders faced both psychological trauma and exposure to particulate matter and toxins during the recovery efforts. The study implies these factors might work synergistically to accelerate aging processes.
“The key takeaway is that PTSD is not only a psychological condition—it is associated with measurable, long-lasting changes in the brain,” Invernizzi said. “In responders exposed to the extreme trauma of 9/11, PTSD was linked to a brain that appears ‘older’ than expected for a person’s chronological age. This underscores the importance of recognizing PTSD as a condition with real biological consequences and reinforces the need for long-term monitoring and support for affected individuals.”
While the findings provide insight into the biological footprint of PTSD, there are limitations to consider. The study utilized a cross-sectional design. This means the data was collected at a single point in time. This structure prevents researchers from proving that the disorder caused the accelerated aging. It remains possible that pre-existing brain differences made some individuals more susceptible to developing the condition.
“It’s important to note that an ‘older-appearing’ brain does not mean inevitable cognitive decline or neurodegenerative disease,” Invernizzi noted. “Brain age is a statistical biomarker, not a clinical diagnosis. Additionally, while our findings show a strong association between PTSD and accelerated brain aging, they do not prove causality.”
Future research efforts will likely focus on longitudinal studies that track participants over many years. Monitoring how these brain age markers change over time could help clarify the direction of the relationship between trauma and aging. Scientists also aim to investigate whether specific treatments or lifestyle interventions might slow or reverse these patterns.
“Our next steps include examining how brain aging relates to cognitive performance, physical health, and functional outcomes over time, as well as identifying factors—such as treatment, resilience, or lifestyle—that may slow or reverse accelerated brain aging in PTSD,” Invernizzi told PsyPost. “Ultimately, we hope this work will inform targeted interventions and improve long-term care for trauma exposed populations.”
“This study also highlights the potential of advanced AI-based neuroimaging tools to capture complex brain changes in real-world clinical populations. By using a data-driven approach trained on thousands of brain scans, we can move closer to objective, scalable biomarkers that complement traditional clinical assessments and help bridge neuroscience and public health.”
The study, “MRI signature of brain age underlying post-traumatic stress disorder in World Trade Center responders,” was authored by Azzurra Invernizzi, Francesco La Rosa, Anna Sather, Elza Rechtman, Ismail Nabeel, R. Sean Morrison, Alison C. Pellecchia, Stephanie Santiago-Michels, Evelyn J. Bromet, Roberto G. Lucchini, Benjamin J. Luft, Sean A. Clouston, Erin S. Beck, Cheuk Y. Tang, and Megan K. Horton.
A recent study suggests that the popular “Sorting Hat Quiz” from the Harry Potter universe may loosely reflect actual personality traits, particularly for fans of the series. The findings indicate that while the quiz captures some real psychological differences, its predictive power relies heavily on the participant’s familiarity with the narrative. These results were published in PLOS One.
Human beings possess a deep-seated drive to engage with storytelling and often identify closely with fictional characters. This tendency frequently manifests in the popularity of online assessments that assign individuals to specific groups within a fictional universe.
The “Sorting Hat Quiz” is a prominent example where users are sorted into one of four Hogwarts Houses based on their responses to situational questions. Prior investigations suggested a correlation between these House assignments and established psychological traits. The authors of the current study sought to verify these associations using more rigorous personality measures. They also aimed to determine if these connections exist for people who are unfamiliar with the books.
“The project actually started in a very down-to-earth way: my coauthors and I are genuine Harry Potter fans, and at some point we found ourselves joking—but also seriously debating—that each of us’ belongs’ to a different Hogwarts House,” said study author Maria Flakus of the Polish Academy of Sciences in Warsaw.
“That naturally led to a more scientific question: is there any real psychological signal behind these identifications, or are they mostly narrative stereotypes and wishful thinking? In other words, we wanted to see whether people’s House alignment (especially the House they feel they are, or want to be) maps onto meaningful differences in their dominant personality characteristics.”
“At the same time, there was a broader gap worth addressing. Sorting-type pop-culture quizzes are massively popular and people often treat the outcomes as surprisingly ‘accurate,’ yet the evidence for whether they track established psychological traits—and under what conditions—is limited and not fully consistent. We were particularly motivated to test whether the Sorting Hat Quiz can tell us something about personality at all, and whether ‘desired’ House membership might be as informative (or even more informative) than the algorithmic assignment—potentially reflecting an ideal self rather than a measured trait profile.”
To examine this, the research team recruited 677 participants through social media platforms. The sample consisted of adults ranging from 18 to 55 years old who were residents of Poland or spoke Polish fluently. The researchers divided the participants into two distinct groups based on their exposure to the series. The first group contained 578 individuals who had read the Harry Potter books. The second group consisted of 99 individuals who had not read the books.
Participants completed the official Sorting Hat Quiz on the Wizarding World website to determine their designated House. They also indicated which House they personally desired to join. To assess personality, the researchers administered the Polish Personality Lexicon, which is based on the HEXACO model. This model measures honesty-humility, emotional stability, extroversion, agreeableness, conscientiousness, and openness to experience.
The study also employed specific scales to measure darker personality aspects known as the Dark Triad. The researchers used the Narcissistic Admiration and Rivalry Questionnaire and the MACH-IV scale (for Machiavellianism). They assessed psychopathy using the Triarchic Psychopathy Measure. Additionally, the Need for Cognition Scale evaluated how much participants enjoyed complex thinking and intellectual challenges.
The data revealed specific patterns among the participants who had read the books. Individuals sorted into Slytherin scored higher on measures of Machiavellianism, narcissism, and psychopathy compared to members of other Houses. These participants displayed traits associated with manipulation and a focus on self-interest. This finding aligns with the fictional portrayal of Slytherin House as ambitious and sometimes cunning.
Participants sorted into Ravenclaw demonstrated a higher need for cognition. This indicates a preference for intellectual engagement and problem-solving activities. This result corresponds well with the Ravenclaw reputation for valuing wit, learning, and wisdom. Those assigned to Gryffindor scored marginally higher on extroversion than the other groups. This suggests a tendency toward social assertiveness and enthusiasm.
Individuals sorted into Hufflepuff reported higher levels of agreeableness and honesty-humility. This aligns with the fictional description of the House as valuing fair play, loyalty, and hard work. However, these participants also reported lower levels of emotional stability. This finding implies a greater tendency to experience worry or a need for emotional support in stressful situations.
“Readers should think of the effects as modest rather than ‘life-defining,'” Flakus told PsyPost. “Even when differences between Houses are statistically reliable, there’s substantial overlap—many people in different Houses look similar on standard trait measures—so House membership explains only a limited share of personality variance. Practically, that means the Sorting Hat result may capture a real tendency at the group level, but it’s not precise enough for individual prediction or decision-making. It’s best viewed as a fun, coarse-grained signal.”
The researchers noted a discrepancy regarding conscientiousness among Hufflepuffs. Previous theories posited that Hufflepuffs would score highest in this trait due to their association with hard work. The current data provided evidence that Hufflepuffs did not score significantly higher in conscientiousness than members of other Houses. This challenges some of the simpler stereotypes associated with the House.
The researchers also analyzed the personality traits of participants based on the House they wanted to join rather than the one they were assigned. The patterns for desired Houses closely mirrored the results for the assigned Houses among readers. For example, those who wished to be in Slytherin scored higher on narcissism and psychopathy. This implies that personal preference is a strong indicator of one’s psychological makeup in this context.
“We were surprised that the pattern of associations pointed not only to traits but also to how people see themselves—self-identification sometimes seemed as informative as the quiz assignment,” Flakus said.
But the relationships between House assignment and personality traits were largely absent in the group of non-readers. While there was a minor link between Gryffindor assignment and extroversion, most other correlations disappeared. The Sorting Hat Quiz failed to predict the “Dark Triad” traits or need for cognition in participants unfamiliar with the books. This suggests that the quiz itself does not function as a standalone personality test.
These findings suggest that the Sorting Hat Quiz is not an effective tool for psychological assessment in a general context. The predictive power of the quiz appears to depend on the participant’s knowledge of the fictional universe. This supports the “narrative collective assimilation hypothesis.” This theory proposes that immersing oneself in a story allows a person to internalize the traits of a specific group within that narrative.
Fans of the series may unconsciously or consciously align their self-perception with the traits of their preferred House. When they answer personality questions, they may do so through the lens of this identity. For non-readers, the questions in the quiz lack this contextual weight. Consequently, their answers do not aggregate into meaningful personality profiles in the same way.
“The key takeaway is that these kinds of pop-culture quizzes can reflect some real personality differences, but they’re not a substitute for validated psychological assessment,” Flakus explained. “Your ‘House’ can be a fun mirror of broad tendencies—and sometimes your preferred House may say as much about your values or ideal self as about your traits—so it’s best used as a playful starting point for self-reflection, not a diagnosis.”
As with all research, there are some limitations to consider. The group of non-readers was relatively small compared to the group of readers. The sample was also predominantly female and recruited via social media. This may affect how well the results represent the general population.
Future inquiries could examine whether these patterns persist across different generations of fans. Researchers might also investigate similar phenomena in other popular fictional universes. Further study is needed to understand how identifying with fictional groups relates to real-world behaviors and values.
“At this point, we don’t have a fixed long-term roadmap yet, but we do see several promising next steps,” Flakus said. “One natural extension would be to test whether similar patterns appear in other pop-culture identity systems—i.e., whether identifying with particular factions, archetypes, or ‘types’ in other franchises relates to established personality traits in comparable ways.”
“We’re also interested in potential generational differences: the Harry Potter universe has a distinct cultural footprint across age cohorts, so it would be valuable to examine whether the mechanisms behind identification (and its links to traits or values) vary by generation.”
“Finally, an important direction is to look more closely at how these quizzes function among people who don’t know the universe at all—in our study we had such a subgroup, but it was small. A larger, more balanced sample would let us more confidently explore whether the quiz captures general psychological tendencies independent of fandom, or whether familiarity and narrative knowledge meaningfully shape the outcomes.”
The new findings regarding the personality structures of Hogwarts Houses align with separate research focused on external economic behaviors. A study published in Small Business Economics by Martin Obschonka and colleagues utilized a massive dataset to examine how these fictional profiles relate to entrepreneurship.
A new study published in Communications Psychology suggests that artificial intelligence systems can be more effective than humans at establishing emotional closeness during deep conversations, provided the human participant believes the AI is a real person. The findings indicate that while individuals can form social bonds with AI, knowing the partner is a machine reduces the feeling of connection.
The rapid development of large language models has fundamentally altered the landscape of human-computer interaction. Previous observations have indicated that these programs can generate content that appears empathetic and similar to human speech. Despite these advancements, it remained unclear whether humans could form relationships with AI that are as strong as those formed with other people. This is particularly relevant during the initial stages of getting to know a stranger.
Scientists aimed to fill this gap by investigating how relationship building differs between human partners and AI partners. They sought to determine if AI could handle “deep talk,” which involves sharing personal feelings and memories, as effectively as it handles superficial “small talk.” Additionally, the research team wanted to understand how a person’s pre-existing attitude toward technology affects this connection. Many people view AI with skepticism or perceive it as a threat to uniquely human qualities like emotion.
To investigate these dynamics, the research team recruited a total of 492 participants between the ages of 18 and 35. The sample consisted of university students. The experiments took place online to mimic typical digital communication. To simulate a realistic environment for relationship building, the researchers utilized a method known as the “Fast Friends Procedure.” This standardized protocol involves two partners asking and answering a series of questions that become increasingly personal over time.
In the first study, 322 participants engaged in a text-based chat. They were all informed that they would be interacting with another human participant. In reality, the researchers assigned half of the participants to chat with a real human. The other half interacted with a fictional character generated by a Google AI model known as PaLM 2. The interactions were further divided into two categories. Some pairs engaged in small talk, discussing casual topics. Others engaged in deep talk, addressing emotionally charged subjects.
The results from this first experiment showed a distinct difference based on the type of conversation. When the interaction involved small talk, participants reported similar levels of closeness regardless of whether their partner was human or AI. However, in the deep talk condition, the AI partner outperformed the human partner. Participants who unknowingly chatted with the AI reported significantly higher feelings of interpersonal closeness than those who chatted with real humans.
To understand why this occurred, the researchers analyzed the linguistic patterns of the chats. They found that the AI produced responses with higher levels of “self-disclosure.” The AI spoke more about emotions, self-related topics, and social processes. This behavior appeared to encourage the human participants to reciprocate. When the AI shared more “personal” details, the humans did the same. This mutual exchange of personal information led to a stronger perceived bond.
The second study sought to determine how the label assigned to the partner influenced these feelings. This phase focused exclusively on deep conversations. The researchers analyzed data from 334 participants, combining new recruits with relevant data from the first experiment. In this setup, the researchers manipulated the information given to the participants. Some were told they were chatting with a human, while others were told they were interacting with an AI.
The researchers found that the label played a significant role in relationship building. Regardless of whether the partner was actually a human or a machine, participants reported feeling less closeness when they believed they were interacting with an AI. This suggests an anti-AI bias that hinders social connection. The researchers noted that this effect was likely due to lower motivation. When people thought they were talking to a machine, they wrote shorter responses and engaged less with the conversation.
Despite this bias, the study showed that relationship building did not disappear entirely. Participants still reported an increase in closeness after chatting with a partner labeled as AI, just to a lesser degree than with a partner labeled as human. This suggests that people can develop social bonds with artificial agents even when they are fully aware of the agent’s non-human nature.
The researchers also explored individual differences in these interactions. They looked at a personality trait called “universalism,” which involves a concern for the welfare of people and nature. The analysis indicated that individuals who scored high on universalism felt closer to partners labeled as human but did not show the same increased closeness toward partners labeled as AI. This finding suggests that personal values may influence how receptive an individual is to forming bonds with technology.
There are several potential misinterpretations and limitations to consider regarding this work. The study relied on text-based communication, which differs significantly from face-to-face or voice-based interactions. The absence of visual and auditory cues might make it easier for an AI to pass as human. Additionally, the sample consisted of university students from a Western cultural context. The findings may not apply to other age groups or cultures.
The AI responses were generated using a specific model available in early 2024. As technology evolves rapidly, newer models might yield different results. It is also important to note that the AI was prompted to act as a specific character. This means the results apply to AI that is designed to mimic human behavior, rather than a generic chatbot assistant.
Future research could investigate whether these effects persist over longer periods. This study looked only at a single, short-term interaction. Scientists could also explore whether using avatars or voice generation changes the dynamic of the relationship. It would be useful to understand if the “uncanny valley” effect, where near-human replicas cause discomfort, becomes relevant as the technology becomes more realistic.
The study has dual implications for society. On one hand, the ability of AI to foster closeness suggests it could be useful in therapeutic settings or for combating loneliness. It could help alleviate the strain on overburdened social and medical services. On the other hand, the fact that AI was most effective when disguised as a human points to significant ethical risks. Malicious actors could use such systems to create deceptive emotional connections for scams or manipulation.
Recent findings in neuroscience provide new evidence that musical creativity is not a static trait but a dynamic process involving the rapid reconfiguration of brain networks. By monitoring the brain activity of skilled jazz pianists, an international research team discovered that high levels of improvisational freedom rely less on introspection and more on sensory and motor engagement. The study suggests that the brain shifts its processing strategy depending on how much creative liberty a musician exerts. These findings were published in the Annals of the New York Academy of Sciences.
Creativity is a complex human ability often defined as the capacity to produce ideas that are both novel and appropriate for a given context. One scientific view proposes that creativity emerges from a balance between constraints and freedom, or between what is predictable and what is surprising. Musical improvisation offers an ideal setting to study this balance because it requires musicians to generate new material spontaneously while adhering to specific structural rules.
Previous neuroimaging studies have identified various brain regions associated with improvisation. These include areas linked to motor planning, emotional processing, and the monitoring of one’s own performance. However, most of these studies have looked at brain activity as a static average over time. This approach can miss the rapid fluctuations in neural connectivity that characterize real-time creative performance. The authors of the current study sought to map these fleeting changes to understand how the brain adapts to different levels of improvisational constraints.
“My main motivation for the study was a long-standing scientific challenge about how to study creativity in real time,” said study author Peter Vuust, the director of the Center for Music in the Brain and professor at Aarhus University and the Royal Academy of Music Aarhus.
“Much research looks at finished products or abstract tasks, but fewer studies capture the process of creating something new as it unfolds in the brain. Musical jazz improvisation offers a rare opportunity because it is spontaneous yet structured—musicians create novel material moment-to-moment while still following certain rules relating to harmony, rhythm and structure.”
“So the gap was twofold: 1) A need for ecologically valid models of creativity (real behavior, not artificial lab tasks). 2) Limited knowledge about how whole-brain networks dynamically reconfigure during different levels of creative freedom.”
“In the Center for Music in the Brain we have the unique capability of studying brain activity as it unfolds in real time, using state-of-the-art brain imaging combined with whole-brain modelling methods which allow for understanding the shifting brain network activity over time,” Vuust explained.
The study included 16 male jazz pianists with significant experience in the genre. All participants were right-handed and had no history of neurological disease. On average, the musicians had over ten years of dedicated jazz practice. The researchers utilized functional magnetic resonance imaging to record brain activity. This imaging technique measures changes in blood flow to infer which areas of the brain are most active.
To allow the musicians to play while inside the MRI scanner, the team used a custom-designed, non-magnetic fiber optic keyboard. This 25-key instrument was positioned on the participants’ laps. This setup allowed the musicians to play with their right hand while listening to audio through noise-canceling headphones.
The experimental procedure involved playing along with a backing track of the jazz standard “Days of Wine and Roses.” The backing track provided the bass and drums to create a realistic musical context. The participants performed under four specific conditions. First, they played the melody of the song from memory. Second, they played an alternate melody from a score sheet they had briefly studied.
The third and fourth conditions introduced improvisation. In the third task, musicians improvised variations based on the melody. In the fourth and final task, they improvised freely based solely on the song’s chord progression. This design created a gradient of creative freedom, ranging from strict memorization to unconstrained expression. Each condition lasted for 45 seconds and was repeated multiple times.
The researchers analyzed the musical output using digital tools to assess complexity. They measured the number of notes played and calculated the “entropy” of the melodies. In this context, entropy refers to the unpredictability of the musical choices. Higher entropy indicates a performance that is less repetitive and harder to predict.
The behavioral results showed the expected relationship between freedom and musical complexity. As the task became less constrained, the musicians played significantly more notes. The condition involving free improvisation on the chord changes resulted in the highest number of notes and the highest level of entropy. The analysis also revealed that during free improvisation, the musicians tended to use smaller intervals between notes. This suggests a dense and rapidly moving musical style.
To analyze the brain imaging data, the researchers employed a method known as Leading Eigenvector Dynamics Analysis. This advanced analytical technique focuses on the phase-locking of blood oxygenation level-dependent signals. It allows scientists to detect recurrent patterns of functional connectivity that may only last for short periods. This is distinct from traditional methods that assume brain connectivity remains constant throughout a task.
The imaging results revealed five distinct brain states, or “substates,” that appeared with varying frequency across the conditions. One of these states was associated with the brain’s reward system. It included the orbitofrontal cortex, a region involved in sensory integration and pleasure. This reward-related state was more active during all playing conditions compared to when the musicians were resting. This finding aligns with the idea that playing music is inherently rewarding, regardless of whether one is improvising or playing from memory.
“A simple takeaway is: Creativity in music is not located in a single ‘creative center’ of the brain,” Vuust told PsyPost. “Instead, it emerges from rapid shifts between multiple brain networks—including those involved in movement, hearing, reward, attention, and self-reflection, depending on the improvisational taks: whether you are trying to improvise on the melody or the chord changes.”
A distinct pattern emerged when the researchers compared the improvisation tasks to the memory tasks. Both the melodic and free improvisation conditions significantly increased the probability of engaging a brain state dominated by auditory and sensorimotor networks, as well as the posterior salience network. These regions are critical for processing sound, coordinating complex movements, and integrating sensory information.
The increased activity in auditory and sensorimotor areas suggests that improvisation places a heavy demand on the brain’s ability to predict and execute sound. Jazz musicians often report “hearing” lines in their head immediately before playing them. The data supports the notion that improvisation is a highly embodied activity. It relies on a tight coupling between the auditory cortex and the motor system to navigate the musical landscape in real time.
Perhaps the most distinct finding appeared in the condition with the highest level of creative freedom. When musicians improvised freely on the chords, the researchers observed a decrease in the occurrence of a brain state involving the default mode network and the executive control network. The default mode network is typically active during introspection, mind-wandering, and self-referential thought. The executive control network is usually involved in planning and goal-directed behavior.
The reduced presence of these networks during free improvisation implies a shift in cognitive strategy. To generate novel ideas rapidly without getting stuck in evaluation or planning, the brain may need to suppress these introspective systems. This aligns with the concept of “flow,” where an individual becomes fully immersed in an activity and self-consciousness recedes. The musicians appeared to rely less on internal planning and more on external sensory feedback.
“Another key message is that greater freedom in improvisation changes how the brain is organized in the moment,” Vuust said. “When musicians improvise more freely, their brains rely more on auditory–motor and salience systems (listening, acting, reacting), and less on heavily controlled, evaluative networks. In everyday terms: creativity often involves letting go of over-analysis while staying highly engaged and responsive.”
The study indicates that creativity involves a flexible reconfiguration of neural resources. Moderate improvisation may require a balance of structure and freedom. However, highly unconstrained improvisation appears to demand a surrender of executive control in favor of sensory-motor processes.
“The effects are not about small local activations but about system-level reconfigurations—which networks are more or less likely to appear over time,” Vuust explained. “Practically, this means the significance lies in patterns and probabilities, not single brain spots lighting up.”
“For musicians and educators, the implication is that training creativity may involve balancing structure and freedom, rather than maximizing one or the other. For neuroscience, it shows that dynamic brain-state analysis can reveal meaningful differences even within subtle variations of the same task.”
As with all research, there are limitations to consider. The sample consisted exclusively of male jazz pianists. This homogeneity limits the ability to generalize the results to female musicians or those from other musical traditions. The creative demands of jazz are specific and may differ from those in other arts, such as painting or writing.
Another consideration is the nature of the “novelty” observed. While the free improvisation condition produced the most unpredictable music, the study did not assess the aesthetic quality of these performances. Higher entropy does not necessarily equate to better music. Previous research suggests that listeners often prefer a balance of complexity and familiarity. The most unconstrained performances might be the most cognitively demanding but not necessarily the most pleasing to an audience.
“Another possible misinterpretation is to assume that more novelty automatically equals more enjoyment or value,” Vuust noted. “The study notes that pleasure and complexity often follow an inverted-U relationship—too much unpredictability can reduce perceived enjoyment.”
Future research could address these gaps by recruiting a more diverse group of participants. Comparing jazz improvisation with other forms of real-time creativity could reveal which brain dynamics are universal and which are specific to music. The authors also suggest that future studies could investigate how these brain states relate to subjective feelings of inspiration or enjoyment. Understanding the link between neural dynamics and the quality of the creative product remains a key goal for the field.
The study, “Creativity in Music: The Brain Dynamics of Jazz Improvisation,” was authored by Patricia Alves Da Mota, Henrique Miguel Fernandes, Ana Teresa Lourenço Queiroga, Eloise Stark, Jakub Vohryzek, Joana Cabral, Ole Adrian Heggli, Nuno Sousa, Gustavo Deco, Morten Kringelbach, and Peter Vuust.
A new study published in the British Journal of Psychology provides evidence that women in the late stages of pregnancy and early motherhood do not display increased submissiveness when facing potential social threats. Contrary to the expectation that physical vulnerability would lead to conflict avoidance, the findings suggest that women in the perinatal period tend to aggressively protect resources when interacting with threatening-looking men.
The rationale behind this investigation is rooted in the evolutionary history of human development. Human infants are born in a state of high dependency, requiring significant time and energy from caregivers to survive. Throughout history, high rates of infant mortality likely necessitated specific cognitive adaptations in parents to help them assess and manage dangers in the environment.
Psychological theories, such as protection motivation theory, propose that people constantly weigh potential threats against their ability to cope with them. When the perceived threat outweighs the ability to cope, individuals typically adopt protective or avoidant behaviors.
This calculation is particularly relevant during pregnancy. The perinatal period, defined as the months leading up to and immediately following childbirth, is physically demanding. Pregnant women experience reduced physical mobility and significant metabolic costs associated with fetal development.
Because of these physical limitations and the high value of the developing fetus, previous models of parental motivation suggested that pregnant women should be highly risk-averse. The logic follows that if a pregnant woman is physically vulnerable, she should avoid escalation and confrontation to prevent harm to herself and her unborn child.
Past research supports the idea that pregnancy heightens sensitivity to danger. For example, pregnant women often show stronger reactions to disgust and are better at recognizing angry or fearful faces than non-pregnant women. The authors of the current study wanted to determine if this heightened sensitivity translates into behavioral submissiveness.
“While previous work demonstrated that pregnancy may change how women perceive threats—such as how fast they spot an angry or fearful face—we didn’t know how this might lead to changes in their actual behavior. Particularly, we became interested in knowing if this enhanced sensitivity to threat may impact their willingness to compete over resources they may need,” said co-author Shawn Geniole, an associate professor at the University of the Fraser Valley.
“On one hand, pregnancy brings new financial and other demands, making it important to compete for and secure resources (e.g., preferred/overtime shifts at work or better products/services). On the other hand, if pregnancy boosts sensitivity to social threats, it may bring greater cautiousness, increasing the likelihood of ‘backing down’ to avoid any risks of conflict or retaliation.”
“We therefore wanted to conduct this study to determine precisely how pregnancy, and more specifically the perinatal period—the months leading up to and immediately after delivery—would impact these types of competitive decisions. To do so, we used an experimental economics task in which women had to decide how to share resources with others.”
The researchers recruited a total of 139 participants. The sample included 86 perinatal women and a control group of 53 non-perinatal women. The perinatal group was tested at two specific time points: approximately 29 weeks into their gestation and again one month after giving birth. The control group also completed testing at two time points separated by a two-month interval to match the timeline of the pregnant participants.
The primary measure used in the study was the “Threat Premium Task.” This is a competitively framed variation of the Ultimatum Game, a standard tool in economic psychology. In this task, participants were given a set amount of virtual money, specifically ten coins, and asked to propose a split with a series of partners. The participants were told that the goal was to keep as much money as possible. However, there was a catch. If the partner accepted the offer, the money was split as proposed. If the partner rejected the offer, neither party received anything.
This design forced participants to make a strategic calculation. Offering a low amount was profitable but risky, as a threatening partner might be perceived as more likely to reject the offer out of spite or aggression. Offering a high amount was safer but resulted in less resource acquisition for the participant. This “threat premium”—the extra money paid to scary-looking partners—is a measure of submissive behavior.
“The women in the study had to carefully balance both the desire to maximize earnings and to avoid retaliation. We were particularly interested in how sensitive they would be—or how much their decisions would change—when interacting with others who appeared more or less threatening.”
The “partners” in this game were not real people but photographs of male faces. Unbeknownst to the participants, these faces had been digitally manipulated to appear either more or less threatening.
The results contradicted the preregistered predictions of the research team. The non-pregnant control group behaved as expected. They were sensitive to the social cues of threat and tended to offer more money to the threatening-looking men than to the non-threatening men. This indicates a typical strategy of appeasement to avoid conflict.
But the perinatal women showed a completely different behavioral pattern. Instead of paying a higher premium to threatening men, they became less generous. The study found that pregnant women were less sensitive to the social threat cues and less willing to cede resources. They dominantly protected their coins rather than submissively handing them over.
This effect was particularly pronounced during the pre-birth session when the women were in the third trimester of pregnancy. The data indicated that the anticipated “threat premium” was effectively eliminated in the perinatal group.
“The biggest takeaway is that pregnancy doesn’t necessarily make women more submissive,” explained co-author Valentina Proietti, an assistant professor at the University of the Fraser Valley. “Based on previous research, we originally expected that pregnant and postpartum women might be more prone to submissive behavior and more likely to relinquish their resources when faced with threatening individuals.”
“However, we found the exact opposite to be true: women in the perinatal period actually defended their resources more dominantly than those who weren’t pregnant, especially when they were dealing with people who looked more threatening. In short: while the common assumption is that heightened threat-sensitivity leads to caution in the face of such threat, our findings suggest it may actually trigger a more dominant drive to secure and protect the resources necessary for themselves and their growing families.”
These findings align with a phenomenon observed in non-human mammals known as maternal aggression. In many species, including rodents and bears, females become significantly more aggressive and protective during pregnancy and lactation. This biological shift prioritizes the security and provision of offspring over the mother’s own safety or tendency toward conflict avoidance.
The researchers suggest that in humans, this maternal defense mechanism may manifest as a refusal to be intimidated by social threats when resources are at stake. The drive to secure necessary assets for the growing family appears to suppress the usual tendency to back down from threatening individuals.
“This pattern may fit with what researchers call ‘maternal aggression’ in other mammals — think of a protective and potentially aggressive mother bear with her cubs,” the researchers told PsyPost. “While we didn’t measure aggression directly, the fact that perinatal women were less submissive in the face of potential threats aligns with this idea.”
“While our study used a more controlled economic task, the results may point toward a more general change in behavior during a truly unique life stage. Readers should think of the perinatal phase as a special/sensitive period—a time when a woman’s social and economic priorities may shift to meet the new demands of motherhood.”
“Although we used a rather simple economic task in our study, the same mechanisms at play here may extend to other types of competitive interactions in the real world, such as bargaining for better work or overtime shifts, navigating online marketplaces, or negotiating for services. We view this study as a first step in understanding how this special biological period reshapes economic decision-making, and we hope to explore these more ‘real-world’ economic interactions in future research.”
The study offers new insights into the psychology of pregnancy, but — as with all research — there are limitations to consider. The study utilized only male faces as the source of social threat.
“Although we’d ideally like to study real‑world economic interactions and other forms of competition that involve a variety of interaction partners, our study focused only on how women responded to threatening situations involving unfamiliar men. As a result, we still don’t know how perinatal women might behave in similar competitive situations with other women. That remains an important direction for future research.”
Additionally, while the sample size was relatively large for this type of research, distinguishing the specific effects of pregnancy from the general effects of parenthood requires even larger groups that compare pregnant women exclusively to women who have never had children.
The study also raises questions about the biological mechanisms driving this behavior. The researchers speculate that hormonal changes may play a key role. Testosterone levels, for instance, are known to rise during pregnancy. In men, higher testosterone is associated with the same type of dominant behavior observed in the perinatal women in this study.
However, the researchers did not measure hormone levels, so this link remains a hypothesis for future investigation. Future work might also explore how this resource-protection drive interacts with the known decreases in mating motivation that occur during pregnancy.
Looking ahead, “we would like to investigate how these effects may extend to real-world economic interactions and how changes in hormones during pregnancy may play a role and/or explain some of the findings here,” Geniole said.
“One ongoing challenge with this kind of research is finding a large enough sample of participants at the right moment in pregnancy or postpartum,” Proietti added. “If you are a professional who supports women during this period—whether you are a midwife, doula, lactation consultant, or work in a maternity ward—and you’d like to see this population be better represented in research, we’d be happy to connect by email at lifespan.lab@ufv.ca or through Instagram (https://www.instagram.com/bicocca_child_and_baby_lab?igsh=dGUxNmdpeDR4djEx) and share information about any future studies! If interested, reader can also check out some additional work at https://bicoccababylab.wixsite.com/website/en.”
A new study suggests that the amount of attention paid to Donald Trump online helps predict optimism on Wall Street. Published in American Behavioral Scientist, the research indicates that spikes in Google searches for the former president tend to precede increases in bullish sentiment among individual investors. This relationship appears to have grown stronger in the period following the 2024 U.S. election.
The financial world has traditionally operated under the assumption that markets are rational. This view holds that asset prices reflect all available information regarding economic fundamentals, such as corporate earnings, interest rates, and employment data. However, the field of behavioral finance challenges this perspective. It argues that human psychology, cognitive biases, and collective emotion play a significant role in how investors make decisions.
Political figures have always influenced markets, but typically this occurs through specific policy decisions or legislative actions. Donald Trump represents a shift in this dynamic. His influence is often exerted through a pervasive media presence and direct communication styles rather than traditional policymaking channels alone. The researchers wanted to understand if the sheer volume of attention a political figure generates can act as a signal for market mood, independent of specific policy details.
“We were motivated by a clear gap between two well-established literatures that rarely talk to each other: behavioral finance has shown that investor sentiment moves markets, and political communication research has shown that media attention shapes public perceptions, but few studies connect political attention directly to financial sentiment,” said study author Raúl Gómez Martínez, an associate professor at Rey Juan Carlos University in Madrid.
“Donald Trump offered a unique real-world case because his media presence is unusually intense and persistent, even outside formal policymaking, raising the question of whether attention alone can influence market psychology. We therefore wanted to test whether high-frequency digital signals, such as Google search activity, could capture this transmission mechanism and help explain weekly changes in retail investor optimism. In short, the study addresses the broader problem of how political narratives spill over into financial markets beyond traditional fundamentals.”
The study builds on the concept that attention is a finite resource. In the digital age, what captures the public’s focus often drives their economic expectations. The researcher sought to test whether the visibility of Donald Trump, a figure closely associated with pro-business narratives, directly impacts investor sentiment. This phenomenon is often referred to by market analysts as the “Trump trade,” where his political prominence signals potential deregulation and tax cuts.
To investigate this connection, the research team analyzed weekly data spanning from April 5, 2020, to October 12, 2025. This timeframe provided a total of 289 weekly observations. The researchers utilized Google Trends to measure public attention. This tool indexes search interest on a scale from zero to 100 rather than providing raw search numbers. It allows for a standardized comparison of interest over time.
The researchers tracked the search term “Donald Trump” within the United States to gauge the intensity of public focus. For investor sentiment, they relied on data from the American Association of Individual Investors (AAII). This non-profit organization conducts a weekly survey asking its members if they feel bullish, bearish, or neutral about the stock market over the next six months. The study focused specifically on the percentage of respondents who reported a bullish or optimistic outlook.
The research team employed statistical models known as ordinary least squares regressions. This method helps identify relationships between the search data and the sentiment survey results. They aimed to see if variations in one variable could explain variations in the other. Additionally, the researchers employed Granger causality tests. This statistical technique helps determine if one time series is useful in forecasting another, effectively checking if changes in attention happen before changes in sentiment.
The analysis revealed a positive association between search volume and investor optimism across the entire five-year period. When online searches for Trump increased, self-reported bullish sentiment among individual investors tended to rise in the same week. The Granger causality analysis provided evidence that the search activity occurred before the shift in sentiment. This suggests that public attention flows into market optimism rather than market optimism driving the search traffic.
The researchers then isolated the data from the period following the 2024 election. This subsample covered the weeks from November 3, 2024, to October 12, 2025. In this smaller set of 50 weeks, the connection between attention and sentiment became much more pronounced. The statistical model explained approximately 15 percent of the variation in investor sentiment during this post-election phase. This is a notable increase compared to about 2 percent for the full five-year period.
The strength of the relationship more than doubled in the post-election data. This indicates that in times of heightened political activity or uncertainty, the market becomes more sensitive to political visibility. The authors suggest that the political context acts as an amplifier. When Trump is at the center of the news cycle during a critical political transition, his visibility becomes a stronger driver of economic expectations for retail investors.
“What we show is that media attention becomes a directly observable, quantifiable variable with real explanatory power for market dynamics,” Gómez Martínez told PsyPost. “Even though the full-sample fit is modest (R²≈0,02), this is common in finance, where sentiment is influenced by many overlapping factors; what matters is that attention consistently adds incremental information.”
“In more sensitive political contexts, the explanatory power rises markedly (R²≈0,15 and a coefficient more than double), indicating that this signal becomes substantially more relevant when uncertainty or polarization is high. In that sense, political attention measured through Google Trends can function as a new complementary market indicator—an additional behavioral barometer that investors and analysts can use alongside traditional economic and financial variables to inform investment decisions.”
These findings imply that financial markets are not driven solely by economic spreadsheets. Collective attention and mass psychology serve as measurable drivers of financial expectations. For the average person, this suggests that everyday news consumption and online behavior can indirectly influence prices by shifting the general mood of investors.
“Our findings show that spikes in public interest in a highly visible political figure like Donald Trump, measured through Google searches, tend to precede increases in investor optimism, meaning that media attention itself can shape market mood,” Gómez Martínez explained. “This suggests that everyday news consumption and online behavior can indirectly influence prices by affecting sentiment, especially among retail investors. In short, politics and digital attention are not just background noise—they can become measurable drivers of financial expectations and market dynamics.”
The study provides a practical application for the theories of behavioral finance. It moves beyond the anecdotal observation that politics moves markets to providing statistical evidence. It supports the idea that high-profile figures can serve as exogenous drivers of sentiment. Their media dominance can shape market psychology even before any concrete policies are enacted.
“Nothing in the results truly surprised us, because they were broadly consistent with what behavioral finance and attention-based theories would predict: highly visible political figures should influence expectations and, therefore, investor sentiment,” Gómez Martínez said. “What was important for us was not discovering an unexpected effect, but demonstrating it rigorously with data, using an econometric framework and supervised regression techniques that allow us to quantify and test the relationship formally.”
“In other words, we moved from an intuitive or anecdotal idea—’politics moves markets’—to statistically validated evidence. That empirical validation is what gives the findings credibility and practical value for both researchers and practitioners.”
While the findings provide evidence of a link between political attention and market mood, the study has certain limitations. The sentiment data comes from the American Association of Individual Investors, which reflects the views of retail investors rather than large institutional firms. Retail investors are often considered more susceptible to behavioral biases and media influence than professional fund managers. It is possible that institutional investors interpret these attention spikes differently.
Google Trends measures the volume of searches but not the intent behind them. A spike in searches could result from negative controversies just as easily as positive news. The tool does not distinguish between a supporter searching for rally dates and a critic searching for details on a scandal. The current study assumes the attention is generally interpreted through the lens of the “Trump trade,” but it does not qualitatively analyze the content of the news driving the searches.
The researchers also note that financial markets are complex ecosystems influenced by countless variables simultaneously. Political attention is one factor among many.
“A potential misinterpretation we would like to preempt is the idea that media attention to a single political figure ‘drives the market’ by itself,” Gómez Martínez told PsyPost. “Our results do not imply that political searches replace fundamentals such as earnings, interest rates, or macroeconomic news; rather, they show that attention adds an additional behavioral layer that helps explain changes in sentiment at the margin. Financial markets are influenced by many overlapping forces, so this variable should be understood as complementary, not deterministic.”
Future research could incorporate sentiment analysis of news headlines. This would allow researchers to determine the tone of the media coverage alongside the volume. Expanding the scope to include institutional investor data would help determine if professional traders react similarly to retail investors. The researchers also suggest applying this methodology to other political figures to see if the phenomenon is unique to Trump.
“This paper is part of an ongoing collaboration between researchers at Universidad Rey Juan Carlos (URJC) and Dublin City University (DCU), and it represents just one step in a broader research agenda focused on understanding investor sentiment as a measurable and actionable variable,” Gómez Martínez explained. “Our long-term goal is to continue developing models that integrate behavioral indicators—such as digital attention, surveys, and online activity—alongside traditional financial data to improve how markets are analyzed and forecasted.”
“We currently have several related articles in progress that expand this line of work using alternative sentiment proxies and more advanced econometric and supervised machine-learning regression techniques to strengthen predictive performance. Importantly, this research also has practical transfer through my fintech, InvestMood, where these insights are applied to build algorithmic trading systems that help investors incorporate sentiment-based signals into real-world investments.”
“Perhaps the most important point to add is that this study illustrates how the growing availability of digital behavioral data is changing the way we can analyze financial markets,” Gómez Martínez said. “Tools such as Google Trends allow us to observe collective attention almost in real time, something that was simply not possible a decade ago, and this opens new opportunities to measure psychological and social drivers of market movements more precisely.”
“More broadly, we hope the paper encourages researchers and practitioners to think beyond purely fundamental variables and to treat attention and sentiment as legitimate, quantifiable components of market dynamics. In that sense, the study is not only about one political figure, but about demonstrating a methodology that can be applied to many other contexts where public narratives influence financial expectations.”
New research conducted in Finland highlights distinct patterns in relationship stability when comparing same-sex and opposite-sex unions. The findings indicate that while female couples experience the highest rates of divorce, the factors contributing to these breakups vary by gender composition. The study suggests that traditional gender norms regarding income and the specific challenges faced by immigrant men in host societies play substantial roles in these outcomes. This research was published in the journal Advances in Life Course Research.
Sociologists and demographers have previously observed that same-sex couples tend to dissolve their unions at higher rates than opposite-sex couples. This trend is particularly pronounced among female couples. Despite this established pattern, the specific reasons behind these disparities have remained largely unexplained. Theoretical models suggest that minority stress, which includes experiences of discrimination and stigma, likely destabilizes these relationships.
Other theories focus on the search for a partner. Finding the right spouse involves predicting future compatibility, a process that is inherently uncertain. This uncertainty is often higher regarding economic characteristics. Researchers wanted to understand if specific observable factors could account for the stability gap. The authors of the current study aimed to determine if nationality intermarriage, religious affiliation, education, or income dynamics could explain the differences in divorce risks.
“There was increasing interest in understanding how the intersections of several minority statuses (e.g., sexual minority and immigration background) shape divorce risks. Not much was known about this because there has been a lack of sufficiently large data to statistically address these types of questions,” said study author Elina Einiö, a lecturer at the Helsinki Institute for Demography and Population Health at the University of Helsinki.
The researchers utilized comprehensive register-based data from Statistics Finland. The dataset covered the entire population of individuals who entered a same-sex registered partnership or an opposite-sex marriage between March 2002 and February 2017. The observation window ended just before Finland implemented gender-neutral marriage laws, replacing the registered partnership system.
The final sample consisted of 3,780 same-sex couples and 339,401 opposite-sex couples. Among the same-sex unions, 37.2 percent were male couples and 62.8 percent were female couples. The researchers restricted the data to couples where at least one spouse lived in Finland at the time of registration and was born in the country. They tracked these couples until the end of 2021 to identify legal divorces.
The analysis employed Cox proportional hazards models to estimate divorce risks. The models controlled for variables such as the year of marriage, the age of both spouses, and the area of residence. The researchers also incorporated annual data on taxable income and religious affiliation based on church tax records.
The general findings revealed a clear hierarchy in divorce risk. Approximately 40 percent of female couples divorced within the first ten years of their legal union. This rate was significantly higher than the 24 percent observed for male couples. Opposite-sex couples had the lowest rate, with 21 percent divorcing within the same timeframe.
For female couples, the elevated risk persisted even after accounting for various socioeconomic factors. The researchers found that income and religious affiliation played only a modest role in explaining their higher divorce rates. The risk for female couples remained roughly double that of opposite-sex couples in the fully adjusted models. This suggests that unobserved factors, potentially including minority stress, continue to impact these relationships heavily.
The results for male couples told a different story. Their slightly higher risk of divorce was partly explained by higher rates of intermarriage and lower rates of religious affiliation. When researchers adjusted for these factors, the difference in divorce risk between male couples and opposite-sex couples became barely significant.
A major focus of the study was the impact of nationality intermarriage. The data showed that marriages involving a foreign-born husband and a native-born spouse were less stable. This pattern was consistent for both male couples and opposite-sex couples. It indicates that the specific experience of being an immigrant man in a host society may strain a marriage.
“It was surprising to see that intermarriage between a foreign-born husband and a native-born spouse destabilizes marriages, regardless of the latter spouse’s gender,” Einiö told PsyPost. “This suggests that there could be psychological distress stemming from being an immigrant man in a host society rather than distress resulting from gendered conflicts between a man and a wife due to different cultural understandings of gender roles.”
This destabilizing effect was not observed in female couples. Marriages between a foreign-born woman and a native-born woman did not show elevated divorce risks compared to couples where both women were native-born.
“Female same-sex couples had an elevated divorce risk, but this risk did not further increase if a native-born woman married a foreign-born wife,” Einiö said. “One of the reasons could be that when a native-born woman legalizes her relationship with another woman, it is often with someone of a relatively similar cultural background (e.g., a wife from another European country).”
Income dynamics provided further insight into how gender norms shape relationship stability. The study distinguished between the primary breadwinner and the secondary breadwinner. In opposite-sex couples, this usually aligned with the husband and wife, respectively. For same-sex couples, the researchers categorized earners based on age to allow for comparison.
High income for the primary breadwinner appeared to stabilize all marriages. This was true regardless of the gender composition of the couple. When the primary earner brought in more money, the risk of divorce decreased across the board.
However, the income of the secondary breadwinner had divergent effects. In opposite-sex marriages, a higher income for the secondary earner was associated with an increased risk of divorce. This aligns with theories regarding the “independence effect,” where financial independence may allow a wife to leave an unhappy marriage.
In contrast, a higher income for the secondary earner in same-sex marriages tended to stabilize the union. This was particularly evident for male couples. The data suggests that male couples benefit from greater income equality within the relationship. While income inequality often protected opposite-sex marriages, it appeared to be a risk factor for same-sex unions.
Religious affiliation also emerged as a significant factor. The study measured this by tracking membership in Finland’s national churches. Joint membership in a church was associated with lower divorce risks for all couple types. This may reflect shared values or the presence of social support from a religious community.
Dissimilarity in religious status was detrimental for some. When one partner was a church member and the other was not, divorce risk increased for same-sex couples. This effect was strongest for male couples. Such dissimilarity did not appear to impact the stability of opposite-sex couples.
The researchers discussed several theoretical implications of these findings. The persistence of high divorce rates among female couples supports the minority stress hypothesis. Women in same-sex relationships may face compounded stress from sexual minority status and gender-related societal expectations. They may also lack the institutional support often available to mixed-gender couples.
The findings regarding men suggest that deviations from cultural norms impact them differently. For immigrant men, the pressure of adapting to a host society appears to bleed into marital stability. For gay men, the lack of shared religious community or significant income disparities can weaken the relationship bond.
The study has some limitations inherent to the use of administrative data. The registers do not contain information on the psychological well-being of the participants. This prevents a direct measurement of relationship quality or specific stressors. The data relies on legal gender markers, which excludes non-binary identities. Additionally, religious affiliation was measured by church tax payment, which may not accurately reflect personal faith or spirituality.
The researchers note that the context of Finland is specific. The country is known for high gender equality but was relatively late among Nordic nations to adopt same-sex marriage laws. The transition from registered partnerships to marriage in 2017 may have altered the social landscape, though the study period largely covers the partnership era.
Future research is needed to see if these patterns hold in other countries. The authors specifically express interest in whether the destabilizing effect of intermarriage for men is consistent across different European nations. Understanding these nuances helps clarify how the intersection of gender, culture, and economic resources influences the longevity of modern relationships.
Login
