Article by David Zendle & Heather Wardle: “Industry data sharing has the potential to revolutionise evidence on video gaming and mental health, as well as a host of other critical topics. However, collaborative data sharing agreements between academics and industry partners may also afford industry enormous power in steering the development of this evidence base. In this paper, we outline how nonfinancial conflicts of interest may emerge when industry share data with academics. We then go on to describe ways in which such conflicts may affect the quality of the evidence base. Finally, we suggest strategies for mitigating this impact and preserving research independence. We focus on the development of data infrastructure: technological, social, and educational architecture that facilitates unfettered and free access to the kinds of high-quality data that industry hold, but without industry involvement…(More)”.
ResearchDataGov
“ResearchDataGov.org is a product of the federal statistical agencies and units, created in response to the Foundations of Evidence-based Policymaking Act of 2018. The site is the single portal for discovery of restricted data in the federal statistical system. The agencies have provided detailed descriptions of each data asset. Users can search for data by topic, agency, and keywords. Questions related to the data should be directed to the owning agency, using the contact information on the page that describes the data. In late 2022, users will be able to apply for access to these data using a single-application process built into ResearchDataGov. ResearchDataGov.org is built by and hosted at ICPSR at the University of Michigan, under contract and guidance from the National Center for Science and Engineering Statistics within the National Science Foundation.
The data described in ResearchDataGov.org are owned by and accessed through the agencies and units of the federal statistical system. Data access is determined by the owning or distributing agency and is limited to specific physical or virtual data enclaves. Even though all data assets are listed in a single inventory, they are not necessarily available for use in the same location(s). Multiple data assets accessed in the same location may not be able to be used together due to disclosure risk and other requirements. Please note the access modality of the data in which you are interested and seek guidance from the owning agency about whether assets can be linked or otherwise used together…(More)”.
A Landscape of Open Science Policies Research
Paper by Alejandra Manco: “This literature review aims to examine the approach given to open science policy in the different studies. The main findings are that the approach given to open science has different aspects: policy framing and its geopolitical aspects are described as an asymmetries replication and epistemic governance tool. The main geopolitical aspects of open science policies described in the literature are the relations between international, regional, and national policies. There are also different components of open science covered in the literature: open data seems much discussed in the works in the English language, while open access is the main component discussed in the Portuguese and Spanish speaking papers. Finally, the relationship between open science policies and the science policy is framed by highlighting the innovation and transparency that open science can bring into it…(More)”
Data Analysis for Social Science: A Friendly and Practical Introduction
Book by Elena Llaudet and Kosuke Imai: “…provides a friendly introduction to the statistical concepts and programming skills needed to conduct and evaluate social scientific studies. Using plain language and assuming no prior knowledge of statistics and coding, the book provides a step-by-step guide to analyzing real-world data with the statistical program R for the purpose of answering a wide range of substantive social science questions. It teaches not only how to perform the analyses but also how to interpret results and identify strengths and limitations. This one-of-a-kind textbook includes supplemental materials to accommodate students with minimal knowledge of math and clearly identifies sections with more advanced material so that readers can skip them if they so choose.
- Analyzes real-world data using the powerful, open-sourced statistical program R, which is free for everyone to use
- Teaches how to measure, predict, and explain quantities of interest based on data
- Shows how to infer population characteristics using survey research, predict outcomes using linear models, and estimate causal effects with and without randomized experiments
- Assumes no prior knowledge of statistics or coding
- Specifically designed to accommodate students with a variety of math backgrounds
- Provides cheatsheets of statistical concepts and R code
- Supporting materials available online, including real-world datasets and the code to analyze them, plus—for instructor use—sample syllabi, sample lecture slides, additional datasets, and additional exercises with solutions…(More)”.
How AI That Powers Chatbots and Search Queries Could Discover New Drugs
Karen Hao at The Wall Street Journal: “In their search for new disease-fighting medicines, drug makers have long employed a laborious trial-and-error process to identify the right compounds. But what if artificial intelligence could predict the makeup of a new drug molecule the way Google figures out what you’re searching for, or email programs anticipate your replies—like “Got it, thanks”?
That’s the aim of a new approach that uses an AI technique known as natural language processing—the same technology that enables OpenAI’s ChatGPT to generate human-like responses—to analyze and synthesize proteins, which are the building blocks of life and of many drugs. The approach exploits the fact that biological codes have something in common with search queries and email texts: Both are represented by a series of letters.
Proteins are made up of dozens to thousands of small chemical subunits known as amino acids, and scientists use special notation to document the sequences. With each amino acid corresponding to a single letter of the alphabet, proteins are represented as long, sentence-like combinations.
Natural language algorithms, which quickly analyze language and predict the next step in a conversation, can also be applied to this biological data to create protein-language models. The models encode what might be called the grammar of proteins—the rules that govern which amino acid combinations yield specific therapeutic properties—to predict the sequences of letters that could become the basis of new drug molecules. As a result, the time required for the early stages of drug discovery could shrink from years to months.
“Nature has provided us with tons of examples of proteins that have been designed exquisitely with a variety of functions,” says Ali Madani, founder of ProFluent Bio, a Berkeley, Calif.-based startup focused on language-based protein design. “We’re learning the blueprint from nature.”…(More)”.
Explore the first Open Science Indicators dataset
Article by Lauren Cadwallader, Lindsay Morton, and Iain Hrynaszkiewicz: “Open Science is on the rise. We can infer as much from the proliferation of Open Access publishing options; the steady upward trend in bioRxiv postings; the periodic rollout of new national, institutional, or funder policies.
But what do we actually know about the day-to-day realities of Open Science practice? What are the norms? How do they vary across different research subject areas and regions? Are Open Science practices shifting over time? Where might the next opportunity lie and where do barriers to adoption persist?
To even begin exploring these questions and others like them we need to establish a shared understanding of how we define and measure Open Science practices. We also need to understand the current state of adoption in order to track progress over time. That’s where the Open Science Indicators project comes in. PLOS conceptualized a framework for measuring Open Science practices according to the FAIR principles, and partnered with DataSeer to develop a set of numerical “indicators” linked to specific Open Science characteristics and behaviors observable in published research articles. Our very first dataset, now available for download at Figshare, focuses on three Open Science practices: data sharing, code sharing, and preprint posting…(More)”.
Smart OCR – Advancing the Use of Artificial Intelligence with Open Data
Article by Parth Jain, Abhinay Mannepalli, Raj Parikh, and Jim Samuel: “Optical character recognition (OCR) is growing at a projected compounded annual growth rate (CAGR) of 16%, and is expected to have a value of 39.7 billion USD by 2030, as estimated by Straits research. There has been a growing interest in OCR technologies over the past decade. Optical character recognition is the technological process for transforming images of typed, handwritten, scanned, or printed texts into machine-encoded and machine-readable texts (Tappert, et al., 1990). OCR can be used with a broad range of image or scan formats – for example, these could be in the form of a scanned document such as a .pdf file, a picture of a piece of paper in .png or .jpeg format, or images with embedded text, such as characters on a coffee cup, title on the cover page of a book, the license number on vehicular plates, and images of code on websites. OCR has proven to be a valuable technological process for tackling the important challenge of transforming non-machine-readable data into machine readable data. This enables the use of natural language processing and computational methods on information-rich data which were previously largely non-processable. Given the broad array of scanned and image documents in open government data and other open data sources, OCR holds tremendous promise for value generation with open data.
Open data has been defined as “being data that is made freely available for open consumption, at no direct cost to the public, which can be efficiently located, filtered, downloaded, processed, shared, and reused without any significant restrictions on associated derivatives, use, and reuse” (Chidipothu et al., 2022). Large segments of open data contain images, visuals, scans, and other non-machine-readable content. The size and complexity associated with the manual analysis of such content is prohibitive. The most efficient way would be to establish standardized processes for transforming documents into their OCR output versions. Such machine-readable text could then be analyzed using a range of NLP methods. Artificial Intelligence (AI) can be viewed as being a “set of technologies that mimic the functions and expressions of human intelligence, specifically cognition and logic” (Samuel, 2021). OCR was one of the earliest AI technologies implemented. The first ever optical reader to identify handwritten numerals was the advanced reading machine “IBM 1287,” presented at the World Fair in New York in 1965 (Mori, et al., 1990). The value of open data is well established – however, the extent of usefulness of open data is dependent on “accessibility, machine readability, quality” and the degree to which data can be processed by using analytical and NLP methods (data.gov, 2022; John, et al., 2022)…(More)”
The Risks of Empowering “Citizen Data Scientists”
Article by Reid Blackman and Tamara Sipes: “New tools are enabling organizations to invite and leverage non-data scientists — say, domain data experts, team members very familiar with the business processes, or heads of various business units — to propel their AI efforts. There are advantages to empowering these internal “citizen data scientists,” but also risks. Organizations considering implementing these tools should take five steps: 1) provide ongoing education, 2) provide visibility into similar use cases throughout the organization, 3) create an expert mentor program, 4) have all projects verified by AI experts, and 5) provide resources for inspiration outside your organization…(More)”.
The rise and fall of peer review
Blog by Adam Mastroianni: “For the last 60 years or so, science has been running an experiment on itself. The experimental design wasn’t great; there was no randomization and no control group. Nobody was in charge, exactly, and nobody was really taking consistent measurements. And yet it was the most massive experiment ever run, and it included every scientist on Earth.
Most of those folks didn’t even realize they were in an experiment. Many of them, including me, weren’t born when the experiment started. If we had noticed what was going on, maybe we would have demanded a basic level of scientific rigor. Maybe nobody objected because the hypothesis seemed so obviously true: science will be better off if we have someone check every paper and reject the ones that don’t pass muster. They called it “peer review.”
This was a massive change. From antiquity to modernity, scientists wrote letters and circulated monographs, and the main barriers stopping them from communicating their findings were the cost of paper, postage, or a printing press, or on rare occasions, the cost of a visit from the Catholic Church. Scientific journals appeared in the 1600s, but they operated more like magazines or newsletters, and their processes of picking articles ranged from “we print whatever we get” to “the editor asks his friend what he thinks” to “the whole society votes.” Sometimes journals couldn’t get enough papers to publish, so editors had to go around begging their friends to submit manuscripts, or fill the space themselves. Scientific publishing remained a hodgepodge for centuries.
(Only one of Einstein’s papers was ever peer-reviewed, by the way, and he was so surprised and upset that he published his paper in a different journal instead.)
That all changed after World War II. Governments poured funding into research, and they convened “peer reviewers” to ensure they weren’t wasting their money on foolish proposals. That funding turned into a deluge of papers, and journals that previously struggled to fill their pages now struggled to pick which articles to print. Reviewing papers before publication, which was “quite rare” until the 1960s, became much more common. Then it became universal.
Now pretty much every journal uses outside experts to vet papers, and papers that don’t please reviewers get rejected. You can still write to your friends about your findings, but hiring committees and grant agencies act as if the only science that exists is the stuff published in peer-reviewed journals. This is the grand experiment we’ve been running for six decades.
The results are in. It failed…(More)”.
The Potentially Adverse Impact of Twitter 2.0 on Scientific and Research Communication
Article by Julia Cohen: “In just over a month after the change in Twitter leadership, there have been significant changes to the social media platform, in its new “Twitter 2.0.” version. For researchers who use Twitter as a primary source of data, including many of the computer scientists at USC’s Information Sciences Institute (ISI), the effects could be debilitating…
Over the years, Twitter has been extremely friendly to researchers, providing and maintaining a robust API (application programming interface) specifically for academic research. The Twitter API for Academic Research allows researchers with specific objectives who are affiliated with an academic institution to gather historical and real-time data sets of tweets, and related metadata, at no cost. Currently, the Twitter API for Academic Research continues to be functional and maintained in Twitter 2.0.
The data obtained from the API provides a means to observe public conversations and understand people’s opinions about societal issues. Luca Luceri, a Postdoctoral Research Associate at ISI called Twitter “a primary platform to observe online discussion tied to political and social issues.” And Twitter touts its API for Academic Research as a way for “academic researchers to use data from the public conversation to study topics as diverse as the conversation on Twitter itself.”
However, if people continue deactivating their Twitter accounts, which appears to be the case, the makeup of the user base will change, with data sets and related studies proportionally affected. This is especially true if the user base evolves in a way that makes it more ideologically homogeneous and less diverse.
According to MIT Technology Review, in the first week after its transition, Twitter may have lost one million users, which translates to a 208% increase in lost accounts. And there’s also the concern that the site could not work as effectively, because of the substantial decrease in the size of the engineering teams. This includes concerns about the durability of the service researchers rely on for data, namely the Twitter API. Jason Baumgartner, founder of Pushshift, a social media data collection, analysis, and archiving platform, said in several recent API requests, his team also saw a significant increase in error rates – in the 25-30% range –when they typically see rates near 1%. Though for now this is anecdotal, it leaves researchers wondering if they will be able to rely on Twitter data for future research.
One example of how the makeup of the less-regulated Twitter 2.0 user base could significantly be altered is if marginalized groups leave Twitter at a higher rate than the general user base, e.g. due to increased hate speech. Keith Burghardt, a Computer Scientist at ISI who studies hate speech online said, “It’s not that an underregulated social media changes people’s opinions, but it just makes people much more vocal. So you will probably see a lot more content that is hateful.” In fact, a study by Montclair State University found that hate speech on Twitter skyrocketed in the week after the acquisition of Twitter….(More)”.