Saturday, October 12, 2024

People with dyslexia and dyscalculia show less bias, study shows

Study is the first to explore how people with the conditions acquire biases developed automatically in the general population

University of Plymouth

Dyslexia and dyscalculia are most commonly acknowledged as posing challenges linked to people’s literacy and numeracy.

However, a new study has shown those with the conditions may have previously unheralded strengths, namely reduced bias against others based on characteristics such as their disability, race or gender.

The research set out to explore whether, and to what extent, those with the two learning difficulties acquire certain traits that most people with no disabilities develop automatically.

To do this, it examined data amassed through Project Implicit’s Implicit Association Test (IAT) database on attitudes to disability.

It analysed both the responses to a number of questions – and the time taken to respond – from more than 460,000 people, including 22,747 with physical disabilities, 14,700 with dyslexia, and 1,721 with dyscalculia,

The results showed those with dyslexia exhibited significantly less implicit bias than those with no disability, something the researchers had predicted based on existing knowledge of the condition and its effects.

However, those with dyscalculia exhibited even less bias, compared to those with dyslexia and from the wider population.

The researchers say the results provide new insights into the development of people with conditions such as dyslexia and dyscalculia.

However, they say they are particularly striking in that they support previous suggestions that not all the impacts of such conditions are as negative as many people might perceive them to be.

Published in the journal Neurodiversity, the study was led by Dr Gray Atherton and Dr Liam Cross from the University of Plymouth’s School of Psychology.

It is part of their wider programme of research looking at the experiences of people with a range of neurodiverse conditions, which has recently involved studies looking at whether board and role-playing games have benefits for those with autism.

Dr Cross said: “For most people, categorising social groups is one of the first things we automatically learn how to do. It means we inherently know things about people’s race and gender and, as a result, begin to exhibit certain implicit and psychological biases. Our study shows there are groups within society who, in line with difficulties developing literacy and numeracy skills, may not develop these biases. Our contention would be that not having these biases is a good thing, for the individuals concerned and for modern society as a whole.”

Dr Atherton added: “Our research always tries to explore if there are ways to consider a learning disability as far more than just an individual or a negative issue. Based on this new study, we can see how people with dyslexia and dyscalculia are potentially not being influenced by many of society’s rules in the way that those without such a condition are. It reinforces the message that not everyone needs to be at the same level of competence, as they can brings other strengths to society’s table.”

Journal

Neurodiversity

DOI

10.1177/27546330241288164

Method of Research

Content analysis

Subject of Research

People

Article Title

People with dyslexia or dyscalculia are less biased: Results of a preregistered study from over 450,000 people on the implicit association test

Article Publication Date

9-Oct-2024

Mcity opens for remote testing of autonomous vehicle technologies, calls for federal standards

Remote testing could accelerate the adoption of autonomous vehicles, but innovators need to know the safety regulations they must meet

University of Michigan

Images/Videos

As Mcity begins welcoming researchers in autonomous and connected vehicle technologies from around the U.S. to be remote users of its physical and virtual testing environment, its leadership is calling for federal standards for safety testing, arguing that the lack of clear goalposts is hampering development.

The opening also coincides with a new industry partnership project announced at the NVIDIA AI Summit.

Following two years of upgrades, the University of Michigan's hub for autonomous testing has officially kicked off its remote program with a series of demonstrations for industry and state and federal government officials. By allowing outside researchers to tap into Mcity research vehicles and testing technologies from afar, U-M seeks to accelerate the development of autonomous vehicle technologies to enable safer and more efficient transportation. Funded by the National Science Foundation, the capabilities are initially available to academic researchers.

"This is particularly valuable for academic institutions, including many that serve minority students, that do not have the resources of large, public institutions like U-M," said Henry Liu, U-M professor of civil and environmental engineering and director of both Mcity and the Center for Connected and Automated Transportation. "We call it Mcity 2.0—the next-generation automated vehicle test facility.

"We believe Mcity 2.0's capabilities will expand the volume and increase the speed of AV research that will be conducted and published. This could help put more refined, safer AVs on our roads sooner."

At the same time, a new Mcity partnership announced at the NVIDIA AI Summit in Washington, D.C., aims to advance its remote testing capabilities for industry as well as academic users. Mcity simulation tools are being tapped for a new project in partnership with NVIDIA and nonprofit research managing firm MITRE. The goal is a full simulation version of Mcity that industry and academic researchers can use to rigorously test and validate AVs, accelerating their development and adoption.

An invitation to researchers across the U.S.

In the coming year, Mcity plans to assist 10 research teams from academic institutions in the U.S. with their testing needs. Researchers can submit proposals to Mcity.

By connecting to Mcity's cloud-based digital infrastructure, researchers operating from their home bases can test their autonomous algorithms in virtual and mixed reality environments. Using 5G wireless communications, they can control physical vehicles and traffic signals on Mcity's test streets, and receive real-time data in return—all while protecting proprietary information.

"We invite researchers from across the country to connect with us and join the effort to accelerate developing and deploying connected and autonomous vehicles that will make transportation safer, cleaner and more accessible," said Karen Thole, the Robert J. Vlasic Dean of Engineering. "Please add your brain power to advancing tomorrow's technologies."

The critical role of federal regulation

While Mcity looks to accelerate that research, U-M researchers see federal leadership—in the form of testing standards and funding—as crucial to keeping it on track. Several unanswered questions lie at the heart of the matter, including:

How should AVs be tested and evaluated? How safe is safe enough?
How do we properly prepare our infrastructure to support AV technologies?
How should AVs be licensed?
How should AVs be insured?

"Typically, regulation follows innovation," Liu said. "With AVs, however, the lack of a federal safety testing framework has created uncertainty that is thwarting innovation and delaying the adoption of technologies that we know can save lives and improve the quality of life for millions of people who do not have reliable access to transportation."

Mcity has developed a two-part methodology for evaluating AV safety, called the Mcity Safety Assessment Program. The first component tests driving algorithms' basic competency in ordinary traffic scenarios. The second component challenges the software with a diverse set of dangerous driving scenarios—representing those that most often result in crashes, injuries and fatalities.

Opened in 2015, Mcity was the world's first purpose built-proving ground for connected and automated vehicles. In 2022, a $5.1 million grant from the National Science Foundation helped U-M develop Mcity's digital infrastructure—leading to Mcity 2.0.

Plant stem cells: Better understanding the biological mechanism of growth control

University of Freiburg

Plants form new leaves, flowers and roots at the tips of shoots and roots, in specific growth regions known as meristems. These meristems contain stem cells that divide as needed and form new cells that develop into specialised tissue. Using the example of plant roots, researchers from Freiburg have now been able to decipher which regulatory mechanisms ensure that growth in the meristem occurs in a controlled manner. The results have been published in the journal Nature Plants.

Stem cells are dependent on signals from other cells

The fact that stem cells can continuously divide and form progenitor cells for specialised tissues is not a matter of course: signals from other cells are necessary to control the properties of stem cells. This dependence on signalling processes is also a protective mechanism. If stem cells were able to multiply uncontrollably, this would lead to uncontrolled growth, as in the case of cancer.

WOX5 is an important signalling molecule that regulates stem cells in the meristem. However, the mechanism through which it does this was previously unknown. A research team led by Prof. Dr. Thomas Laux, a member of the CIBSS – Centre for Integrative Biological Signalling Studies Cluster of Excellence at the University of Freiburg, has now succeeded in decoding this mechanism. The team identified HAN as an essential factor that transmits the function of WOX5.

The gene-regulating molecule HAN is an important regulator for plant growth.

“We were able to show that HAN transmits the WOX5 signal and ensures that the CDF4 gene remains inactive in stem cells,” explains Laux. “CDF4 would otherwise cause stem cell properties to be inhibited. By suppressing CDF4, HAN allows the stem cells in the root meristem to remain undifferentiated and continue dividing.”

The team used molecular biology methods as well as mathematical modelling. These provide a possible explanation as to why the seemingly complicated mechanism could be an advantage for the plant: the involvement of HAN as a link between WOX5 and CDF4 appears to make the regulation of stem cells less sensitive to environmental influences. “In further investigations, we now want to find out whether the multi-level nature of the process actually has the effect that we see in the modelling,” says Laux.

Process also important for plant breeding

A precise understanding of the processes by which plants grow is an important basis for breeding more resilient or higher-yielding crops. This is because it allows the targeted identification and selection of plants that can grow and produce yields even under less-than-ideal conditions, such as extreme weather.

DOI

10.1038/s41477-024-01810-z

Using AI and iNaturalist, scientists build one of the highest resolution maps yet of California plants

Citizen scientists can help monitor the state's plants in the face of global change

University of California - Berkeley

California plant distributions from iNaturalist data using AI — image:
Using more than 652,000 observations uploaded to iNaturalist (left), UC Berkeley scientists created an AI model to predict the distribution of 2,221 species of plants around the state. To train the deep learning network, the species observations were linked to 256 x 256 meter remote sensing images taken from the 2012 National Agricultural Imagery Program (right) and with climate variables.
view more
Credit: Moi Exposito-Alonso and Lauren Gillespie, UC Berkeley

With the help of deep learning, University of California, Berkeley, scientists have leveraged citizen science data from the app iNaturalist to create the highest resolution maps yet of plant distributions throughout California.

iNaturalist is a widely-used cellphone app, originally developed by UC Berkeley students, that allows people to upload photos and the location data of plants, animals or any other life they encounter and then crowdsource their identity. The app currently has more than 8 million users worldwide who collectively have uploaded more than 200 million observations.

The researchers used a type of artificial intelligence called a convolutional neural network, which is a deep learning model, to correlate the citizen science data for plants in California with high-resolution remote-sensing satellite or airplane images of the state. The network discovered correlations that were then used to predict the current range of 2,221 plant species throughout California, down to scales of a few square meters.

Botanists usually build high-quality maps of species by painstakingly listing all plant species in an area, but this is not feasible outside of a few small natural areas or national parks. Instead, the AI model, called Deepbiosphere, leverages free data from iNaturalist and remote sensing airplanes or satellites that now cover the entire globe. Given enough observations by citizen scientists, the model could be deployed in countries lacking detailed scientific data on plant distributions and habitats to monitor vegetation change, such as deforestation or regrowth after wildfires.

The findings were published Sept. 5 in the journal Proceedings of the National Academy of Sciences by Moisés "Moi" Expósito-Alonso, a UC Berkeley assistant professor of integrative biology, first author Lauren Gillespie, a doctoral student in computer science at Stanford University, and their colleagues. Gillespie currently has a Fulbright U.S. Student Program grant to use similar techniques to detect patterns of plant biodiversity in Brazil.

“During my year here in Brazil, we’ve seen the worst drought on record and one of the worst fire seasons on record,” Gillespie said. “Remote sensing data so far has been able to tell us where these fires have happened or where the drought is worst, and with the help of deep learning approaches like Deepbiosphere, soon it will tell us what’s happening to individual species on the ground.”

"That is a goal — to expand it to many places," Expósito-Alonso said. "Almost everybody in the world has smartphones now, so maybe people will start taking pictures of natural habitats and this will be able to be done globally. At some point, this is going to allow us to have layers in Google Maps showing where all the species are, so we can protect them. That's our dream."

Apart from being free and covering most of Earth, remote sensing data are also more fine-grained and more frequently updated than other information sources, such as regional climate maps, which often have a resolution of a few kilometers. Using citizen science data with remote sensing images — just the basic infrared maps that provide only a picture and the temperature — could allow daily monitoring of landscape changes that are hard to track.

Such monitoring can help conservationists discover hotspots of change or pinpoint species-rich areas in need of protection.

"With remote sensing, almost every few days there are new pictures of Earth with 1 meter resolution," Expósito-Alonso said. "These now allow us to potentially track in real time shifts in distributions of plants, shifts in distributions of ecosystems. If people are deforesting remote places in the Amazon, now they cannot get away with it — it gets flagged through this prediction network."

Expósito-Alonso, who moved from Stanford to UC Berkeley earlier this year, is an evolutionary biologist interested in how plants evolve genetically to adapt to climate change.

"I felt an urge to have a scalable method to know where plants are and how they're shifting," he said. "We already know that they're trying to migrate to cooler areas, that they're trying to adapt to the environment that they're facing now. The core part of our lab is understanding those shifts and those impacts and whether plants will evolve to adapt."

In the study, the researchers tested Deepbiosphere by excluding some iNaturalist data from the AI training set and then later asking the AI model to predict the plants in the excluded area. The AI model had an accuracy of 89% in identifying the presence of species, compared to 27% for previous methods. They also pitted it against other models developed to predict where plants are growing around California and how they will migrate with rising temperatures and changing rainfall. One of these models is Maxent, developed at the American Museum of Natural History, that uses climate grids and georeferenced plant data. Deepbiosphere performed significantly better than Maxent.

They also tested Deepbiosphere against detailed plant maps created for some of the state's parks. It predicted with 81.4% accuracy the location of redwoods in Redwood National Park in Northern California and accurately captured (with R2=0.53) the burn severity caused by the 2013 Rim Fire in Yosemite National Park.

"What was incredible about this model that Lauren came up with is that you are just training it with publicly available data that people keep uploading with their phones, but you can extract enough information to be able to create nicely defined maps at high resolution," Expósito-Alonso said. "The next question, once we understand the geographic impacts, is, “Are plants going to adapt?’"

Megan Ruffley, also of the Carnegie Institution for Science at Stanford, is a co-author of the paper. The research, which began when Expósito-Alonso was a staff associate at the Carnegie Institution for Science and an assistant professor at Stanford, was funded by the Carnegie Institution for Science.