William Marcellino

Lies, Misinformation Play Key Role in Israel-Hamas Fight

At the strategic level, the war in Gaza is about perception, telling a story about who is the victim and who is the aggressor. Lies, mistruths, and disinformation will play a key and continuing role in this fight.

The United States Isn't Ready for the New Age of AI-Fueled Disinformation

Generative AI offers China and other bad actors the potential to fulfill longstanding desires to shape the global conversation. It is vital that the U.S. government and social media platforms recognize this threat and work together to address it immediately.

AI, ChatGPT, and Language as Technology: Q&A with William Marcellino

William Marcellino, a senior behavioral and social scientist at RAND and professor at the Pardee RAND Graduate School, discusses the rapidly expanding reach of artificial intelligence, the challenges it could pose for both society and policymakers, and how the research community is poised to help.

The Will to Fight in the Age of Social Media

Social media messaging has played a decisive role in strengthening Ukraine's will to fight—arguably the single most important factor in war—against Russia.

Reining in COVID-19 Disinformation from China, Russia, and Elsewhere

During the pandemic, misinformation and conspiracy theories have spread more virulently than ever before. The vast scale of the problem means scalable solutions like machine learning could be needed to rein in the bots, trolls, and conspiracy theories being spread by bad-faith actors.

The Link Between Extremism and Military Functioning

U.S. Department of Defense policy prohibits service members from actively participating in extremist activities. Broadening the policy to include passive forms of participation could introduce two challenges.

A Machine Learning Approach Could Help Counter Disinformation

Disinformation has become a central feature of the COVID-19 crisis. This type of malign information and high-tech “deepfake” imagery poses a risk to democratic societies worldwide by increasing public mistrust in governments and public authorities. New research highlights new ways to detect and dispel disinformation online.

'Star Trek' at 50: How the TV Series Inspired a Boy to Become a Scientist

William Marcellino discusses how “Star Trek” convinced him at an early age that science and the advancement of human knowledge could make the world a better place.

Commentary

The Rise of Generative AI and the Coming Era of Social Media Manipulation 3.0: Next-Generation Chinese Astroturfing and Coping with Ubiquitous AI

This Perspective explores the implications that generative artificial intelligence (AI) might have for social media manipulation. The authors argue that generative AI presents a revolutionary improvement for the social media manipulation process.

Report

Planning Ethical Influence Operations: A Framework for Defense Information Professionals

Using the principles of necessity, effectiveness, and proportionality, the authors develop a framework for evaluating influence operations with an initial screening, conducting a full ethical risk assessment, and preparing a justification statement.

Report

The U.S. Air Force Diversity, Equity, and Inclusion Values on Social Media: A Proof-of-Concept Examination of How Air Force–Affiliated Accounts Represent the Service Online

This report presents a proof-of-concept to assess the online behavior of U.S. Air Force (USAF)-affiliated users over social media, specifically in regard to USAF diversity, equity, and inclusion policies.

Commentary

A House Reunited: Prospects for Bipartisanship in a Divided Country

Based on workshop discussions, this Perspective presents some proposed solutions, and potential data sources to evaluate these solutions, to problems related to mis- and disinformation, election security, extremism, and immigration reform.

Research Brief

A Dangerous Web: Mapping Racially and Ethnically Motivated Violent Extremism

The authors reviewed literature on racially or ethnically motivated violent extremism (REMVE), analyzed social media data from six platforms, and created a global network map to better understand how to inform counter-REMVE strategies online.

Report

Mapping White Identity Terrorism and Racially or Ethnically Motivated Violent Extremism: A Social Network Analysis of Online Activity

The authors reviewed literature on racially or ethnically motivated violent extremism (REMVE), analyzed social media data from six platforms, and created a global network map to better understand how to inform counter-REMVE strategies online.

Draft

Bag-of-Words Algorithms Can Supplement Transformer Sequence Classification & Improve Model Interpretability

Describes a simple method of improving performance on the problem of classifying sequences of text by concatenating the hidden state of a BERT-based transformer model with a dictionary-based bag-of-words model.

Journal Article

Conspiracy Theories Have Much in Common. Their Differences May Hold the Key to Identifying When They Turn Violent.

RAND/Google collaboration to improve machine learning models to detect and understand the rhetorical functions of conspiracy theories over social media.

Tool

Command Narrative Smart Guide

This quick reference provides the big-picture basics for successful command narrative development and deployment.

Report

Developing, Disseminating, and Assessing Command Narrative: Anchoring Command Efforts on a Coherent Story

Drawing on the scientific literature and lessons learned from the joint force, the authors of this report describe best practices for the effective use of command narrative.

Report

Bad Actors in News Reporting: Tracking News Manipulation by State Actors

Researchers found that both Russia and China appear to have manipulated information during the pandemic in service to their respective global agendas. This second report in a series describes these efforts and the workflows used to uncover them.

Commentary

Natural Language Processing: Security- and Defense-Related Lessons Learned

This reference document presents a collection of lessons learned by practitioners from RAND Corporation defense and security projects that employed natural language processing tools and methods.

Report

Rapid Analysis of Foreign Malign Information on COVID-19 in the Indo-Pacific: A Proof-of-Concept Study

The authors explore whether news and blog data could provide some initial insights on the reach of, content of, and tactical strategies used in foreign malign information about COVID-19 in the Indo-Pacific.

Report

Insights from the Bin Laden Archive: Inventory of research and knowledge and initial assessment and characterisation of the Bin Laden Archive

The authors present an inventory of current knowledge on Al Qa'ida and of completed and ongoing research on the Bin Laden Archive. They also describe their initial assessment and characterisation of the Bin Laden Archive.

Report

Detecting Conspiracy Theories on Social Media: Improving Machine Learning to Detect and Understand Online Conspiracy Theories

Using linguistic and rhetorical theory, researchers developed an improved model of machine-learning technology to detect conspiracy theory language. This report describes the results and suggests ways to counter effects of online conspiracy theories.

Report

Army Analytic Capabilities: A Case Study Within Army Contracting Command and Its Implications

The U.S. Army lacks access to modern, commonplace analytical data management tools, and its data are locked away in siloed and proprietary databases. To solve these two challenges, the authors found an effective, efficient, and quick path forward.

Draft

Resonance Analysis: A Methodology for Detecting Author Affiliations through Distinctive Patterns of Word Choice

Describe resonance analysis, an evolving text mining method for estimating how much affinity exists in a broader population for a specific group, and provides detailed guidance for using it effectively.

Report

Counter-Radicalization Bot Research: Using Social Bots to Fight Violent Extremism

As the online recruitment of violent extremist organizations grows, the U.S. government may benefit from promising emerging technology tools to rapidly detect targets of such recruitment efforts and deliver counter-radicalization content to them.

Report

Foreign Interference in the 2020 Election: Tools for Detecting Online Election Interference

This report, the second in a series on information efforts by foreign actors, describes what appears to be an instance of election interference on Twitter and offers recommendations for response.

Report

Human-machine detection of online-based malign information

RAND Europe developed a method to detect the malign use of information online for the UK Ministry of Defence's Defence and Security Accelerator.

Report

Detecting Malign or Subversive Information Efforts over Social Media: Scalable Analytics for Early Warning

The authors examine scalable analytic means to detect malign or subversive information efforts. The report has operational relevance to the U.S. government and may also benefit other researchers and social media companies dealing with malign actors.

Research Brief

Will to Fight: Returning to the Human Fundamentals of War

This brief argues for the importance of improving our understanding of what influences will to fight, and it describes two models for doing so: one for analyzing a military unit's will to fight, and one for analyzing a nation's will to fight.

Research Brief

Facts Versus Opinions: How the Style and Language of News Presentation Is Changing in the Digital Age

This brief describes results of a study focused on changes in news presentation over time and across platforms, finding a general shift from objective event-based reporting to reporting that relies more heavily on argumentation and advocacy.

Report

News in a Digital Age: Comparing the Presentation of News Information over Time and Across Media Platforms

This report presents a quantitative assessment of how news presentation has changed over time and how it varies across platforms from more-objective event- and context-based reporting to more-subjective reporting that includes more emotional appeals.

Research Brief

Evaluating Los Angeles County's Mental Health Community Engagement Campaign

In order to evaluate how well a Los Angeles County mental health campaign met its goals, RAND conducted an in-person survey of event attendees, a social media analysis, and a web-based survey of a broader population of Los Angeles youth.

Report

Evaluation of Los Angeles County's Mental Health Community Engagement Campaign

To evaluate how well a Los Angeles County mental health campaign met its goals, RAND conducted an in-person survey of event attendees, a social media analysis, and a web-based survey of a broader population of Los Angeles youth.

Journal Article

Social Media in Africa: A Double-Edged Sword for Security and Development: Research Brief

The United Nations Development Programme commissioned RAND Europe to explore social media use and its links to online radicalisation in Africa. The study analysed the social media use of three Islamist terrorist groups across seven countries.

Journal Article

Social Media in Africa: A Double-Edged Sword for Security and Development

The United Nations Development Programme commissioned RAND Europe to explore social media use and its links to online radicalisation in Africa. The study analysed the social media use of three Islamist terrorist groups across seven countries.

Journal Article

Social Media in Africa: A Double-Edged Sword for Security and Development

The United Nations Development Programme commissioned RAND Europe to explore social media use and its links to online radicalisation in Africa. The study analysed the social media use of three Islamist terrorist groups across seven countries.

Journal Article

Social Media in Africa: A Double-Edged Sword for Security and Development: Technical Annex

An annex of supporting information for the report, containing thematic content in relation to the countries examined and the three case study terrorist groups analysed, alongside content relating to the Twitter data analysis and interviews conducted.

Report

Will to Fight: Analyzing, Modeling, and Simulating the Will to Fight of Military Units

This report defines and describes will to fight and provides a model of unit will to fight that can be applied to ground combat units of any scale. It also provides a theoretical basis for adding will to fight to military war gaming.

Report

Russian Social Media Influence: Understanding Russian Propaganda in Eastern Europe

Researchers analyzed Russian social media data and conducted interviews with regional and security experts to understand the critical ingredients to countering Russia's propaganda campaign against former Soviet states.

Report

Air Force Sexual Assault Situations, Settings, and Offender Behaviors

To assist Air Force efforts to prevent and respond to sexual assault, this report focuses on providing a better understanding of sexual assaults committed by airmen, including suspect characteristics and behaviors and risky situations and settings.

Report

Monitoring Social Media: Lessons for Future Department of Defense Social Media Analysis in Support of Information Operations

To support the U.S. Department of Defense in expanding its capacity for social media analysis, this report reviews the analytic approaches that will be most valuable for information operations and considerations for implementation.

Report

Dominating Duffer's Domain: Lessons for the U.S. Marine Corps Information Operations Practitioner

To support the U.S. Marine Corps and its curriculum for information operations personnel, RAND has adapted the premise of the military classic The Defence of Duffer's Drift for a modern-day audience of Marine Corps officers.

Report

Dominating Duffer's Domain: Lessons for the U.S. Army Information Operations Practitioner

Adapting the format of military classic, The Defence of Duffer's Drift, this report analyzes alternative courses of action in the same hypothetical mission in order to derive lessons about integrating information operations into operational planning.

Journal Article

Detecting Changes in Newspaper Reporting of Suicide After a Statewide Social Marketing Campaign

In conclusion and similar to many past evaluations, we find no evidence that the KTS campaign was effective in changing the way in which suicides are reported on in California had any effect.

Report

Considerations for Integrating Women into Closed Occupations in U.S. Special Operations Forces

This report assesses challenges for unit cohesion from integrating women into special operations forces and provides analytical support for validating occupational standards for positions controlled by U.S. Special Operations Command.

Report

Hazing in the U.S. Armed Forces: Recommendations for Hazing Prevention Policy and Practice

This report addresses ways to improve the armed forces' definition of hazing, the effects of and motivations for hazing, how the armed forces can prevent and respond to hazing, and how the armed forces can improve the tracking of hazing incidents.

Journal Article

Revisioning Strategic Communication Through Rhetoric and Discourse Analysis

Strategic communication can be made more conceptually robust and draw on a more powerful and useful suite of tools and methods by borrowing from two language-focused disciplines: rhetoric and discourse analysis.

Journal Article

Talk Like a Marine: USMC Linguistic Acculturation and Civil–military Argument

This study examines the relationship between US Marine discourse and civil–military public argument.

Journal Article

"I Don't Think I Would Have Recovered": A Personal and Sociocultural Study of Resilience Among U. S. Marines

USMC resilience training should focus on skills and concepts needed to resolve challenges to values in the lives of Marines. Technical-medical interventions should be reserved for clinical populations.