Articles tagged with Intelligence Testing
Researchers found that biological neural systems are more efficient in learning with limited samples, outperforming deep reinforcement learning algorithms in a Pong simulation. This breakthrough suggests actual intelligence may be biological.
A study led by Penn State's Sara Dommer found that people cheat on tasks like crossword puzzles and calorie tracking when intrinsic rewards are at stake. Participants reported higher scores and attributed their performance to intelligence, rather than cheating, even after being graded objectively.
Researchers developed DeepLens design method based on curriculum learning to optimize complex lens designs. The approach considers key parameters like resolution, aperture, and field of view, providing optimal solutions without human intervention.
Researchers found that adults prioritize cognitive cues over vocal signals when evaluating a child's intelligence or negative emotions, while vocal cues dominate for assessing happiness or helplessness. This study provides insights into the complex dynamics of caregiver-child interactions and their evolutionary roots.
Researchers at Mayo Clinic used AI to analyze electroencephalogram (EEG) tests, identifying subtle brain wave patterns characteristic of cognitive problems like Alzheimer's disease. This technology has the potential to provide healthcare professionals with a more accessible tool for early diagnosis in communities without easy access to...
A recent study published in iScience found that newborn chicks are predisposed to learn quicker based on the colour of their mother. The research team discovered a colour-dependent bias in learning speed, with chicks exposed to a deep blue mother learning faster than those raised by mothers with other colours.
Researchers found that AI-generated submissions went undetected in a UK university's examinations system, earning higher grades than real students' answers. The study suggests that AI tools could potentially allow students to cheat and achieve better grades than their peers.
Researchers found associations between childhood cognition, adolescent crystalized intelligence, and midlife fluid intelligence with DNA methylation age acceleration. A 1-unit increase in adolescent PPVT was associated with a 0.048-year decrease in aging.
A study found that neural balance, measured by the E/I ratio, is associated with brain maturity and better cognitive ability. Research suggests that as children develop, their E/I ratios decrease, leading to improved performance in cognitive tests.
A new machine-learning model is providing teachers with access to high-quality, fit-for-purpose creativity tests that can score assessments in a fraction of the time and cost. The algorithm marks a test in a single millisecond, saving teachers thousands of hours.
Researchers from six teams in five labs worldwide used self-driving labs to discover 21 top-performing OSL gain candidates, accelerating the discovery process by months. The decentralized workflow enabled rapid replication of experimental findings and democratized the discovery process.
A study by Georgia State University found that people rate AI responses as more virtuous and intelligent than human responses on moral issues. The AI system, ChatGPT, was able to fool participants into believing its responses were better, highlighting the need for a deeper understanding of AI's role in society.
A new study found that GPT-4 and Google Gemini performed poorly in breast imaging classification, with a high percentage of discordant assignments that could impact patient management. The researchers emphasize the need for regulation of these large language models in high-level medical scenarios.
A culturally-responsive training intervention based on science of reading and math improved comprehension, visualization, and connection-making skills among 66 third-grade English language learners. The program enhanced students' performance in math by utilizing direct and explicit teaching methods, focusing on key cognitive concepts.
A study found that when adult male blue tits were absent, young males had higher success rates in fathering offspring outside of their breeding pair. Young males' poor performance was due to competition with older males, not lack of experience. In the absence of adult males, 33% of young males fathered at least one extra-pair offspring.
Researchers at Eötvös Loránd University discovered that dogs have a 'canine g factor' similar to humans, influencing cognitive abilities and ageing patterns. The study found a global cognitive decline in dogs over two and a half years, with poorer health affecting the rate of decline.
A study by Dr Lixin Ren found that extracurricular activities have a limited impact on math skills in preschool children, with effects dependent on behavioral regulation and approaches to learning. For children with poor self-regulation skills, structured activities can improve early math skills.
A new study published in eClinicalMedicine suggests that ECG-AI can flag some risks years sooner than current risk calculator equations by identifying signs of coronary artery disease, such as calcification and blockages. The technology has the potential to save more lives by identifying people who do not know they have coronary disease.
The ISOOS system uses a smart carrier as the mobile base and control center, connecting multiple UAVs, USVs, and AUVs for simultaneous 3D observation of targeted regions. This enables fast mapping of ocean floor and studying submesoscale eddies, previously difficult to access due to variability in space and time.
A study published in the Journal of Pediatrics found that feeding infants formula with a complex component of milk (MFGM) and lactoferrin for 12 months raised IQ by 5 points at 5 ½ years of age, improving speed of processing information and visual-spatial skills. The effects were seen long after the end of formula feeding.
A WVU team is developing technology to detect and track lethal non-trackable space debris. The project aims to identify objects that can cause mission-ending damage to satellites and other space assets. The researchers will use a digital twin and simulation environment to test and evaluate their findings.
A new study published in Pediatrics suggests that pediatric concussions do not impact intelligence or IQ levels. The research, which combined data from two cohort studies in Canada and the US, found no significant difference in IQ scores between children with concussion and those with orthopedic injuries.
Researchers found that both higher cognitive abilities and exploratory behavior are linked to longer life expectancy in wild gray mouse lemurs. Animals with better cognitive performance and physical condition tended to have longer lifespans.
A study led by Dartmouth College researchers found that participants' color memory improved when stimuli were meaningful and part of real-world objects. The results demonstrate that the capacity for visual working memory of colors is more continuous and flexible than previously thought, with better recall of colors in meaningful contexts.
Researchers found that ChatGPT 4 can accurately identify appropriate imaging services for breast cancer screening and breast pain. The AI model outperformed its predecessor, ChatGPT 3.5, in providing evidence-backed guidelines for radiologists.
A study found ChatGPT answered less than half of the test questions correctly, scoring 46% initially and 56% after a month. The AI tool performed well on general medicine questions but struggled with ophthalmology subspecialties.
A new study published in Nature finds that artificial intelligence (AI) surpasses human sonographers in accurately assessing cardiac function from echocardiogram studies. The research, conducted at the Smidt Heart Institute, demonstrates AI's potential to improve the quality and effectiveness of heart evaluations.
A study by University of Cologne researchers found that virtual reality gamers completed tasks faster with higher levels of general intelligence and processing capacity. The results suggest VR games can be used as a tool for predicting job performance, bridging the gap between research and practice.
Researchers developed AI models based on UNet and MobileNet architectures to analyze standardized abnormalities in CT images, accurately identifying object presence and confidence. These models achieved an absolute percentage error of less than 5 percent, comparable to human professionals.
After six months of disrupted schooling during the COVID-19 pandemic, German students scored substantially lower on intelligence tests compared to earlier cohorts. The gap in test scores persisted even after sixteen months of irregular schooling.
Researchers explore how AI language models like ChatGPT understand and respond to user input, mirroring their users' intelligence. The Reverse Turing Test reveals that chatbots reflect the intelligence level of their interviewers, incorporating their biases into responses.
A study published in PLOS Digital Health found that ChatGPT scored between 52.4% and 75.0% on the USMLE exams, exceeding the passing threshold of approximately 60%. The model demonstrated high concordance rates and produced significant insights for most responses.
Researchers developed AI algorithms to predict ICU delirium risk, showing promise in identifying patients at high risk. The models can detect early warning signs and provide continuously updated risk predictions, enabling clinicians to devote more resources to interventions.
A large Danish population study found that premature birth before 34 weeks is linked to lower scores in maths and language tests as a teenager. However, the researchers noted that cognitive outcomes are not predetermined at birth but are heavily influenced by social circumstances.
A new study has mapped the parts of the brain that support fluid intelligence, a key feature of human cognition. The research found that patients with damage to the right frontal regions performed poorly in tasks requiring problem-solving and reasoning.
A recent study found that AI failed to pass a radiology qualifying examination, with an average accuracy of 79.5%. However, the researchers suggest that further training and revision could lead to improved results.
A new analysis offers the clearest picture yet of how various brain regions and neural networks contribute to a person's problem-solving ability. The network neuroscience theory, which posits that intelligence emerges from the global architecture of the brain, outperformed other theories in predicting general intelligence.
Researchers developed an AI model to analyze spatial and temporal gait parameters, identifying key features for diagnosing Parkinson's. The model achieved high diagnostic accuracy, reducing the probability of error in clinical assessments.
A new AI-powered blood test, DELFI, has been developed to detect liver cancer with high accuracy. The test detected over 80% of liver cancers in a study of 724 individuals, with an overall sensitivity of 88% and specificity of 98%.
Researchers developed a method combining musical tests with EEG measurement to detect cognitive decline in old age. The study tested 50 elderly people who scored low on the mini-mental test, showing promising results in detecting mental deterioration.
Researchers from Martin Luther University Halle-Wittenberg found that people with the Impostor Phenomenon devalue their objectively measured performance and attribute positive results to external causes. The study showed no relation between actual intelligence or performance and self-reported degree of Impostor Phenomenon.
A Swedish study of over 9,000 children found that playing video games increased their intelligence by 2.5 IQ points compared to TV-watching or social media. The study controlled for genetic differences and socioeconomic background, suggesting a positive effect of video games on cognitive abilities.
A new study from the University of Kansas found that creativity assessments in education rely on standardized approaches, overlooking many students. The research highlights the need for a more diverse range of assessments to identify and encourage creative young people across different domains and populations.
Researchers at West Virginia University are using machine learning and geographic information systems to identify areas with low COVID-vaccine uptake. They aim to pinpoint counties with increased risk of outbreaks, predict where testing is most crucial, and develop targeted interventions to increase testing rates. By acknowledging comm...
A study found that UK adults' accuracy in estimating national statistics is influenced by their intelligence, religiousness, and political beliefs. Respondents were most accurate when estimating characteristics like vegetarianism and car ownership, but least accurate for minority groups like homosexuals.
Researchers from South Ural State University created a hybrid intelligence system that can predict pile bearing capacity with high accuracy. The system, developed by combining adaptive neuro-fuzzy inference system and group method of data handling, outperformed traditional methods and provided close to measured values.
China's space transportation systems have made significant leaps in recent decades, with advancements in launch vehicles, propulsion systems, and artificial intelligence. The country aims to become a powerful space nation by the mid-21st century, with plans for manned missions to the Moon and Mars.
Research found that COVID-19 patients with more severe symptoms scored lower on reasoning and problem-solving tasks. Cognitive deficits were greatest in those receiving mechanical ventilation, similar to average decline seen over a decade of ageing.
The study used a slap on the hand as a test subject to examine reflexes and their coordination. The researchers found that humans anticipate pain when reacting to potential harm, which could be applied to robot development for improved safety. Robots can learn from human reflex systems to develop intelligent protective mechanisms.
Research at the University of Strathclyde found that people with high emotional intelligence are more likely to identify fake news correctly. Participants who scored highly in emotional intelligence tests made fewer errors and were more discerning in their assessment of veracity.
The E-Morph test uses artificial intelligence to identify substances that can have oestrogen-like effects on breast cells. This method has the potential to replace animal experiments currently required for detecting hormone-like effects.
A study published by the University of Edinburgh found that exposure to air pollution in childhood is associated with a decline in thinking skills up to 60 years later. The research used historical data to estimate exposure levels and analyzed the relationship between air pollution and cognitive change throughout life.
A new study using chess game data shows that cognitive capacity increases steadily until around the middle of the fourth decade before plateauing. Modern 30-year-olds exhibit a higher level of cognitive ability than those born 100 years ago at the same age.
A 12-week exercise regimen improved physical fitness in active-duty airmen, with added benefits from consuming a nutrient-enriched beverage. The study found significant gains in strength, endurance, mobility, and stability, as well as enhanced cognitive function, including better episodic memory and fluid intelligence.
Researchers developed a single factor, 'p', combining different measures of psychiatric outcomes, which predicts patient prognosis and need for extra support. A high 'p' value is linked to adverse outcomes, such as suicide risk, criminality, or drug abuse.
A brief entrance test can predict students' success within their first year of study in Economics, outperforming intelligence tests and school grades. The WiWiSET project found that the test assesses domain-specific knowledge and skills, allowing for a reliable prognosis of study progress.
A new coronavirus app coupled with machine intelligence will enable individuals to get an at-home risk assessment based on their symptoms and travel history. The app will also provide real-time information to local and public health officials, aiding in the targeting of prevention and treatment initiatives.
A new study by University of Pennsylvania researchers found that people, smokers and non-smokers alike, were often inaccurate in their answers to questions about smoking's health effects. However, when participants discussed their ideas with others, especially in a networked setting, they became more accurate in their assessments.
Researchers are developing a framework to assess the 'intelligence' of AI systems by grading them on problem-solving skills and adaptability. The AIQ test will evaluate systems based on accuracy, time taken, and data requirements.
A comprehensive meta-analysis of over 160 studies involving 42,000 students from 27 countries reveals that emotionally intelligent students tend to achieve better academic performance. Higher emotional intelligence is linked to improved grades, achievement test scores, and social relationships with teachers, peers, and family.