Brittany I. Davidson
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Brittany I. Davidson has a total of 165 co-authors. Here are the top co-authors by number of shared publications:
- David A. Ellis
University of Bath
10 shared publications - Heather Shaw
Lancaster University
10 shared publications David Alexander Ellis
9 shared publications- Douglas Anderson Parry
Vrije Universiteit Amsterdam Faculty of Social Sciences and Humanities Amsterdam, the
7 shared publications - Jacob Taylor Fisher
Michigan State University
5 shared publications - Jade Davies
University College London, UK
5 shared publications
Latest Publications
- Open Access
- 29 October 2026
- Published by Center for Open Science in PsyArXiv
From motivation to experience: Blurred boundaries in the measurement of adolescent social media engagement
Despite extensive research on the role of social media use in adolescent mental health, fundamental questions remain about what current self-report measures capture. We systematically reviewed 96 adolescent self-report measures of social media use, comprising 1,526 unique items. Through content analysis, we examined what measures assess and the behavioural and psychological processes through which they are operationalised. We found substantial conceptual overlap between measures, particularly those assessing problematic use and addiction, alongside discrepancies between item content and young people’s reported experiences. Importantly, current measures were found to frequently combine motivation, behaviour, experience, and mindset, despite these psychological behaviours and processes being theoretically distinguished. This raises questions about whether existing measures meaningfully test the hypothesises mechanisms and theoretical frameworks underpinning social media and mental health research. Our findings highlight the need to move beyond content-focused approaches to measurement, towards greater attention to how social media is operationalised and recognition of its potentially dynamic and interdependent nature.
- 29 October 2026
- Published by Center for Open Science in PsyArXiv
Open AccessFrom motivation to experience: Blurred boundaries in the measurement of adolescent social media engagement
Despite extensive research on the role of social media use in adolescent mental health, fundamental questions remain about what current self-report measures capture. We systematically reviewed 96 adolescent self-report measures of social media use, comprising 1,526 unique items. Through content analysis, we examined what measures assess and the behavioural and psychological processes through which they are operationalised. We found substantial conceptual overlap between measures, particularly those assessing problematic use and addiction, alongside discrepancies between item content and young people’s reported experiences. Importantly, current measures were found to frequently combine motivation, behaviour, experience, and mindset, despite these psychological behaviours and processes being theoretically distinguished. This raises questions about whether existing measures meaningfully test the hypothesises mechanisms and theoretical frameworks underpinning social media and mental health research. Our findings highlight the need to move beyond content-focused approaches to measurement, towards greater attention to how social media is operationalised and recognition of its potentially dynamic and interdependent nature.
- Open Access
- 12 October 2026
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 58 (16) , 1-37
Sentiment Analysis in Digital Spaces: An Overview of Reviews
Digital data generated via social media have become a prosperous entity for sentiment analysis researchers seeking to understand individuals’ feelings, attitudes, and emotions. Numerous systematic reviews have synthesized work across diverse contexts, media, methods, and applications; however, they rarely address the validity of sentiment analysis methods or critically examine scientific practices. Our overview of 66 reviews comprises 3,131 unique primary studies using sentiment analysis to dissect online digital data. We provide a high-level overview of current applications, methods, outcomes, and common challenges in sentiment analysis research. A bespoke risk of reporting bias (RoRB) framework was designed to assess the transparency, completeness, and consistency of reporting practices within the included systematic reviews, with particular attention to methodological clarity and adherence to reporting standards. We found diverse applications, methods, outcomes, and persistent challenges outlined in the reviews, which we discuss in relation to the validity of sentiment analysis research. Importantly, reporting practices across the reviews were limited or inconsistent, a concern we examine by considering how existing systematic reviews shape understanding of the field and influence subsequent decisions by researchers and practitioners. We also outline how future research can address these issues and highlight their importance across numerous societal applications.
- 12 October 2026
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 58 (16) , 1-37
Open AccessSentiment Analysis in Digital Spaces: An Overview of Reviews
Digital data generated via social media have become a prosperous entity for sentiment analysis researchers seeking to understand individuals’ feelings, attitudes, and emotions. Numerous systematic reviews have synthesized work across diverse contexts, media, methods, and applications; however, they rarely address the validity of sentiment analysis methods or critically examine scientific practices. Our overview of 66 reviews comprises 3,131 unique primary studies using sentiment analysis to dissect online digital data. We provide a high-level overview of current applications, methods, outcomes, and common challenges in sentiment analysis research. A bespoke risk of reporting bias (RoRB) framework was designed to assess the transparency, completeness, and consistency of reporting practices within the included systematic reviews, with particular attention to methodological clarity and adherence to reporting standards. We found diverse applications, methods, outcomes, and persistent challenges outlined in the reviews, which we discuss in relation to the validity of sentiment analysis research. Importantly, reporting practices across the reviews were limited or inconsistent, a concern we examine by considering how existing systematic reviews shape understanding of the field and influence subsequent decisions by researchers and practitioners. We also outline how future research can address these issues and highlight their importance across numerous societal applications.
- Open Access
- 5 October 2026
- Published by Center for Open Science in PsyArXiv
Towards Natural Behaviour Processing for Representation and Prediction of Digital Human Behaviour
Studying how people behave and understanding why their actions unfold differently aremajor goals of psychological science. However, most research relies on nomotheticapproaches that aggregate behavioural data into summary statistics, obscuringperson-specific patterns that afford person-specific understanding. The progressingproliferation of digital technologies such as smartphones, wearables, and computermediatedsystems now generates fine-grained, temporally ordered records of everydaybehaviour, creating unprecedented opportunities to study both systematic interindividualdifferences and patterns of intra-individual variation. Here, we proposeNatural Behaviour Processing (NBP) as a framework for analysing digital behaviouraldata using sequence modelling approaches adapted from natural languageprocessing. We argue that behaviour shares critical structural properties withlanguage: both involve a repository of possible actions (vocabulary), systematicpatterns governing how elements typically combine (grammar), and context-dependentmeaning (semantics). These structural parallels motivate applying sequence-basedmodels that have revolutionised language processing to the study of behaviour morebroadly. We review pioneering studies on diverse topics of research, including lifeoutcomes, decision-making, social media use, consumer behaviour, online education,and health, demonstrating that sequential approaches capture person-specific patternsthat are typically lost in aggregation. We discuss the challenges facing NBP adoption,including the tension between predictive modelling and theoretical understanding, theneed for large-scale curated behaviour datasets, and methodological trainingrequirements. The adoption of NBP could transform the way psychologyoperationalises, conceptualises, and ultimately understands human behaviour bysimultaneously accounting for person-specificity in real-world complexity whilemaintaining predictive performance at population levels.
- 5 October 2026
- Published by Center for Open Science in PsyArXiv
Open AccessTowards Natural Behaviour Processing for Representation and Prediction of Digital Human Behaviour
Studying how people behave and understanding why their actions unfold differently aremajor goals of psychological science. However, most research relies on nomotheticapproaches that aggregate behavioural data into summary statistics, obscuringperson-specific patterns that afford person-specific understanding. The progressingproliferation of digital technologies such as smartphones, wearables, and computermediatedsystems now generates fine-grained, temporally ordered records of everydaybehaviour, creating unprecedented opportunities to study both systematic interindividualdifferences and patterns of intra-individual variation. Here, we proposeNatural Behaviour Processing (NBP) as a framework for analysing digital behaviouraldata using sequence modelling approaches adapted from natural languageprocessing. We argue that behaviour shares critical structural properties withlanguage: both involve a repository of possible actions (vocabulary), systematicpatterns governing how elements typically combine (grammar), and context-dependentmeaning (semantics). These structural parallels motivate applying sequence-basedmodels that have revolutionised language processing to the study of behaviour morebroadly. We review pioneering studies on diverse topics of research, including lifeoutcomes, decision-making, social media use, consumer behaviour, online education,and health, demonstrating that sequential approaches capture person-specific patternsthat are typically lost in aggregation. We discuss the challenges facing NBP adoption,including the tension between predictive modelling and theoretical understanding, theneed for large-scale curated behaviour datasets, and methodological trainingrequirements. The adoption of NBP could transform the way psychologyoperationalises, conceptualises, and ultimately understands human behaviour bysimultaneously accounting for person-specificity in real-world complexity whilemaintaining predictive performance at population levels.
- Open Access
- 25 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 2416-2433
The Accountability Paradox: How Platform API Restrictions Undermine AI Transparency Mandates
- 25 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 2416-2433
Open AccessThe Accountability Paradox: How Platform API Restrictions Undermine AI Transparency Mandates
- Open Access
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-20
Good Fences Make Good Neighbors: What's Acceptable (and What's Not) in Neighborhood Video Surveillance
- 13 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-20
Open AccessGood Fences Make Good Neighbors: What's Acceptable (and What's Not) in Neighborhood Video Surveillance
- Open Access
- 3 April 2026
- Published by Wiley in British Journal of Social Psychology
- Vol. 65 (2) , e70079
Indirect mobilisation and violence legitimation through influencers on alternative platforms
Existing mobilisation literature has largely focused on groups and collective sensemaking processes as the primary drivers of collective action. However, online influencers have emerged as key leaders and mobilisers, which can shape collective action through one‐to‐many communication. Using self‐categorisation theory, we examine indirect mobilisation and the legitimation of violence during the August 2024 UK riots through a case study of a far‐right influencer's Telegram channel. The dataset consists of 230 posts and 156 pieces of multimedia content from the Tommy Robinson News Telegram channel from 29th July to 7th August 2024. We employed an abductive thematic analysis approach, revealing how throughout the progression of the riots, posts in the Telegram channel construct group identities, establish epistemic authority and leadership legitimacy, and legitimise violence. Our findings extend the social identity approach of mobilisation into the digital realm, revealing how broadcast‐style, unidirectional affordances of Telegram channels can impact the dynamics of leadership, identity construction and mobilisation of (violent) collective action.
- 3 April 2026
- Published by Wiley in British Journal of Social Psychology
- Vol. 65 (2) , e70079
Open AccessIndirect mobilisation and violence legitimation through influencers on alternative platforms
Existing mobilisation literature has largely focused on groups and collective sensemaking processes as the primary drivers of collective action. However, online influencers have emerged as key leaders and mobilisers, which can shape collective action through one‐to‐many communication. Using self‐categorisation theory, we examine indirect mobilisation and the legitimation of violence during the August 2024 UK riots through a case study of a far‐right influencer's Telegram channel. The dataset consists of 230 posts and 156 pieces of multimedia content from the Tommy Robinson News Telegram channel from 29th July to 7th August 2024. We employed an abductive thematic analysis approach, revealing how throughout the progression of the riots, posts in the Telegram channel construct group identities, establish epistemic authority and leadership legitimacy, and legitimise violence. Our findings extend the social identity approach of mobilisation into the digital realm, revealing how broadcast‐style, unidirectional affordances of Telegram channels can impact the dynamics of leadership, identity construction and mobilisation of (violent) collective action.
- Open Access
- 27 March 2026
- Published by Elsevier in Computers in Human Behavior Reports
- Vol. 22
Understanding the key domains of adolescent social media experience: a parallel Delphi study with young people and key adults
- 27 March 2026
- Published by Elsevier in Computers in Human Behavior Reports
- Vol. 22
Open AccessUnderstanding the key domains of adolescent social media experience: a parallel Delphi study with young people and key adults
- Open Access
- 15 January 2026
- Published by Center for Open Science in PsyArXiv
Understanding the key domains of adolescent social media experience: a parallel Delphi study with young people and key adults
This study aimed to identify the key domains of adolescent social media experience that relate to mental health, and how perspectives on these differ between adolescents and adults. Whilst there has been an explosion in studies looking at the relationship between social media use and mental health, little attention has been paid to the measurement of social media experiences. To fill this gap, we conducted a three-round, online parallel Delphi study with two panels: (1) an adult panel made up of five stakeholder groups (n=39); and (2) an adolescent panel made up of individuals aged 11-15 (n=27). Round 1 used open-text responses to generate 44 statements about adolescent social media experiences, which participants rated their agreement with over two subsequent rounds via iterative, anonymous voting. 18 statements reached consensus, covering ten domains of experience. Statements that reached consensus for both panels related to (1) peer pressure, (2) anxiety, (3) exclusion, (4) body image, (5) self-esteem, and (6) addiction. Statements that reached consensus for the adult panel related to (7) relationships, and (8) validation, and statements that reached consensus for the adolescent panel related to (9) coping/emotion regulation, and (10) empowerment. These findings contribute to a more balanced conceptualisation of social media experience and will guide the development of a psychometric measure. Further research is needed to explore the nuances in adolescent and adult perspectives, including regarding relational experiences, how adolescents feel they can influence others, and adolescents’ conceptualisation of social media as a coping mechanism.
- 15 January 2026
- Published by Center for Open Science in PsyArXiv
Open AccessUnderstanding the key domains of adolescent social media experience: a parallel Delphi study with young people and key adults
This study aimed to identify the key domains of adolescent social media experience that relate to mental health, and how perspectives on these differ between adolescents and adults. Whilst there has been an explosion in studies looking at the relationship between social media use and mental health, little attention has been paid to the measurement of social media experiences. To fill this gap, we conducted a three-round, online parallel Delphi study with two panels: (1) an adult panel made up of five stakeholder groups (n=39); and (2) an adolescent panel made up of individuals aged 11-15 (n=27). Round 1 used open-text responses to generate 44 statements about adolescent social media experiences, which participants rated their agreement with over two subsequent rounds via iterative, anonymous voting. 18 statements reached consensus, covering ten domains of experience. Statements that reached consensus for both panels related to (1) peer pressure, (2) anxiety, (3) exclusion, (4) body image, (5) self-esteem, and (6) addiction. Statements that reached consensus for the adult panel related to (7) relationships, and (8) validation, and statements that reached consensus for the adolescent panel related to (9) coping/emotion regulation, and (10) empowerment. These findings contribute to a more balanced conceptualisation of social media experience and will guide the development of a psychometric measure. Further research is needed to explore the nuances in adolescent and adult perspectives, including regarding relational experiences, how adolescents feel they can influence others, and adolescents’ conceptualisation of social media as a coping mechanism.
- Open Access
- 30 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-14
Using Community Notes to Enact Attitudinal Change towards Urban Wildlife
- 30 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-14
Open AccessUsing Community Notes to Enact Attitudinal Change towards Urban Wildlife
- Open Access
- 26 June 2025
- Published by Center for Open Science in PsyArXiv
Disambiguating sentiment annotation: Using mixed methods to understand annotator experience and impact of instructions on annotation quality
Human-annotated datasets are central to the development and evaluation of sentiment analysis and other text classification systems. However, many existing datasets suffer from low annotator agreement and errors, raising concerns about the quality of the data that computational systems should learn from or align with. Improving annotation quality demands close examination of how datasets are created and how annotators interpret and approach the task. To this end, we create AmbiSent, a new sentiment dataset designed to capture cases of interpretive complexity that commonly challenge both annotators and computational models. Using a mixed-methods approach, we investigate how annotation instructions influence label consistency and annotator experience. Two groups of crowdworkers annotated the dataset under either minimal or detailed instructions, allowing comparison of inter- and intra-annotator agreement and annotation rationales. Our findings reveal that detailed instructions alone do not ensure more consistent annotations – either across or within individuals. Their effectiveness appears contingent on participants’ level of task engagement and the extent to which the instructions align with intuitive annotation strategies. Despite the presence of detailed guidance, participants often defaulted to simplifying the task. However, we also observed sentence types where detailed instructions may improve annotation quality, (e.g., sentences with perspective-dependent sentiment, rhetorical questions, and sentences containing sarcasm). A thematic analysis of open-ended responses further contextualised these findings, offering insights into the annotators’ cognitive effort involved and the practical challenges faced. Together, these results inform recommendations for enhancing task engagement and instruction adherence, offering practical insights for future dataset development. Finally, to support diverse use cases, we release three versions of the AmbiSent dataset, each accompanied by detailed annotator information and label distributions to better accommodate different user needs.
- 26 June 2025
- Published by Center for Open Science in PsyArXiv
Open AccessDisambiguating sentiment annotation: Using mixed methods to understand annotator experience and impact of instructions on annotation quality
Human-annotated datasets are central to the development and evaluation of sentiment analysis and other text classification systems. However, many existing datasets suffer from low annotator agreement and errors, raising concerns about the quality of the data that computational systems should learn from or align with. Improving annotation quality demands close examination of how datasets are created and how annotators interpret and approach the task. To this end, we create AmbiSent, a new sentiment dataset designed to capture cases of interpretive complexity that commonly challenge both annotators and computational models. Using a mixed-methods approach, we investigate how annotation instructions influence label consistency and annotator experience. Two groups of crowdworkers annotated the dataset under either minimal or detailed instructions, allowing comparison of inter- and intra-annotator agreement and annotation rationales. Our findings reveal that detailed instructions alone do not ensure more consistent annotations – either across or within individuals. Their effectiveness appears contingent on participants’ level of task engagement and the extent to which the instructions align with intuitive annotation strategies. Despite the presence of detailed guidance, participants often defaulted to simplifying the task. However, we also observed sentence types where detailed instructions may improve annotation quality, (e.g., sentences with perspective-dependent sentiment, rhetorical questions, and sentences containing sarcasm). A thematic analysis of open-ended responses further contextualised these findings, offering insights into the annotators’ cognitive effort involved and the practical challenges faced. Together, these results inform recommendations for enhancing task engagement and instruction adherence, offering practical insights for future dataset development. Finally, to support diverse use cases, we release three versions of the AmbiSent dataset, each accompanied by detailed annotator information and label distributions to better accommodate different user needs.