Introducing Mediated Statement Analysis (Msa): Examining Truthful or Untruthful Social Media Posts to Decode Deception Online

Arnold, Stewart, and Richard (2024) developed Mediated Statement Analysis (MSA), a novel methodology adapting forensic Statement Analysis techniques for detecting deception in social media posts. Their pilot study analyzed 1,414 truthful and untruthful social media posts using nine categories of textual and linguistic cues, with specific application to COVID-19 misinformation.

Exploratory pilot study employing mixed-methods analysis of 1,414 pre-classified truthful/untruthful social media posts, with detailed examination of a COVID-19 subset (n=130) across three MSA categories adapted from traditional forensic Statement Analysis.

Background & Context

The rise of misinformation in mediated communication has created an urgent need for effective methods to detect deceptive content online, particularly as traditional deception detection relies heavily on nonverbal cues that are absent in digital contexts. The pervasive nature of social media has fundamentally transformed how deception occurs and spreads, creating new challenges for researchers and practitioners alike.

Statement Analysis (SA), a method commonly utilized in law enforcement to assess veracity, has traditionally focused on face-to-face interrogation contexts but required modification for application to mediated communication research. This study represents a pioneering effort to adapt forensic linguistic techniques originally developed for in-person contexts to the unique constraints and characteristics of social media platforms, where brevity, informality, and asynchronous communication dominate.

The research addresses a critical gap in digital deception detection methodology during an era when misinformation—particularly related to health crises like COVID-19—can have serious public health and safety consequences. By developing a systematic framework specifically designed for analyzing text-based social media posts, the researchers aim to provide both theoretical advancement and practical tools for identifying potentially deceptive online content.

Research Design & Methodology

The researchers developed Mediated Statement Analysis (MSA) as a systematic research methodology comprising nine (9) categories of textual and linguistic cues designed to assist with detecting potentially deceptive social media posts online. This framework was specifically engineered to address the unique characteristics of digital communication that differ substantially from spoken statements.

The pilot study analyzed a dataset of 1,414 truthful-untruthful social media posts. The researchers employed a mixed-methods approach that combined quantitative content analysis with qualitative linguistic pattern identification. Posts were pre-classified as truthful or deceptive, allowing the team to identify systematic differences in textual and linguistic markers between the two categories.

Key methodological elements included:

• Development of nine distinct MSA categories adapted from traditional Statement Analysis techniques
• Analysis of textual and linguistic cues including word choice, grammatical structures, and rhetorical patterns
• Application of MSA to a COVID-19 pandemic subset (n = 130) analyzing patterns across three (3) MSA categories
• Systematic comparison of linguistic features between verified truthful and deceptive social media posts

The research had three primary goals: (1) uncovering observations about mediated deception in social media posts, (2) testing and modifying current SA methodology for mediated communication contexts, and (3) developing a textual deception detection methodology applicable to mediated communication. The exploratory nature of the research allowed for iterative refinement of the MSA framework based on empirical findings.

Results & Key Findings

The research successfully established that textual and linguistic cues can differentiate truthful from deceptive social media posts using a systematic analytical framework. The development of the nine-category MSA system represents a significant methodological advancement for digital deception detection.

Specific findings from the analysis revealed:

• Distinctive patterns emerged across the nine MSA categories when comparing truthful versus untruthful posts
• The pilot dataset of 1,414 posts yielded a COVID-19 subset (n = 130) that demonstrated clear applications of MSA for analyzing health-related misinformation
• Textual and linguistic markers proved identifiable and analyzable in the constrained, informal text format typical of social media
• The adapted methodology successfully bridged the gap between traditional forensic Statement Analysis and modern digital communication contexts

The application of MSA to social media posts about COVID-19 proved particularly relevant given the crisis of misinformation online related to health communication narratives and lived experiences during the global pandemic. This demonstrated the methodology's practical utility for addressing real-world problems where deceptive health information can have serious consequences.

The findings suggest that despite the absence of vocal tone, facial expressions, and other nonverbal cues available in face-to-face deception detection, written language in social media posts contains sufficient systematic variation to permit discrimination between truthful and deceptive content when analyzed through the appropriate framework.

Discussion & Significance

This research represents a foundational contribution to the emerging field of digital deception detection by successfully adapting established forensic linguistic methods to contemporary social media contexts. The modification of traditional Statement Analysis methodology for mediated communication contexts addresses a critical methodological gap, providing researchers and practitioners with a theoretically grounded approach to analyzing online veracity.

The significance extends beyond academic advancement. The findings have particular relevance for the creation and distribution of crisis messages, as the dissemination of misinformation in crisis events may be particularly critical. During public health emergencies, natural disasters, or security threats, the ability to rapidly identify deceptive information could support more effective crisis communication strategies and public safety interventions.

The nine-category MSA framework provides a more systematic and potentially trainable approach compared to ad-hoc content moderation or purely automated detection systems. By grounding the methodology in established Statement Analysis principles while adapting it for digital contexts, the researchers have created a bridge between forensic science and communication research that could inform both fields. The work also contributes to the broader understanding of how deception manifests differently across communication modalities.

Limitations & Considerations

As an exploratory pilot study, this research has inherent limitations that should inform interpretation and application of findings. The dataset of 1,414 posts, while substantial for initial exploration, may not capture the full diversity of deceptive tactics employed across different social media platforms, topics, or user demographics. The COVID-19 subset was notably smaller at 130 posts, which may limit the generalizability of findings specific to health misinformation.

Additional considerations include:

• The study does not appear to report inter-rater reliability metrics for classification of posts as truthful versus deceptive, which is critical for establishing methodological rigor
• Platform-specific communication norms (Twitter character limits, Facebook post conventions, etc.) may influence linguistic patterns independent of veracity
• The nine MSA categories are described but detailed operational definitions and coding procedures require further validation
• Cultural and linguistic variation across different user populations may affect the applicability of identified patterns
• The methodology requires human expert analysis rather than automated detection, which may limit scalability for large-scale applications

Practical Applications

The MSA framework has immediate applicability for professionals working in crisis communication, public health messaging, and digital media literacy. Organizations concerned with misinformation—from social media platforms to government health agencies—could employ MSA-trained analysts to identify potentially deceptive content requiring fact-checking or additional scrutiny. The systematic nature of the nine-category framework makes it amenable to training programs for content moderators, journalists, and public information officers.

For law enforcement and investigative contexts, MSA extends the utility of traditional Statement Analysis into digital evidence examination. Investigators analyzing social media posts related to criminal cases, threats, or fraudulent schemes could apply MSA principles to assess credibility of online statements. The methodology may also support development of semi-automated detection tools that flag posts with linguistic features associated with deception for human review, combining the efficiency of technology with the nuanced judgment of trained analysts.