Stream Two
Digital Archives
Becoming a Digital Historian: A Database of Women’s Applications for Re-naturalisation in Britain, 1915-1923
Charlotte Clare - Keele University
In this lightning talk I will discuss my journey into the digital humanities, and the challenges I’ve faced along the way. This will include the process of attempting to teach myself the coding language Python, and then trying to adapt these new skills to create a workable database of women who applied for readmission to British nationality in 1915-1923, the physical copies of which are held at The National Archives at Kew. I will also demonstrate one of the digital methods I have used, text mining, in order to highlight how using technology like this can add to what we know about citizenship, identity and belonging in the early twentieth century.
Making Sense of Public Records Data
Joshua Hurwitz - University of Liverpool
Democratic governments around the world have instituted transparency laws that allow members of the public to request records of State proceedings. In some respects, these records resemble other archival data familiar to social scientists. However, there are some stark differences, as well. These include the means that researchers access these records, as well as unique considerations regarding the completeness and bias contained in these records. In this talk, I will explore how researchers can employ open records laws as a unique and useful data source, as well as the pitfalls and challenges therein. I employ the case of U.S. state-level open records laws, and the particular case of the process of constructing lockdown exemptions in the early COVID-19 pandemic, as a case to illustrate some important points regarding how these laws can be employed in social science, and particularly, organisational, research.
Silence in Court: What is Not Said.
Claire Jones - University of Liverpool
Paul Grice’s theory of Conversational Implicature (1975) has been a cornerstone in the discipline of pragmatics within linguistics for several decades. Whilst Grice’s theory offers a systematic account explaining how people use and understand language non-literally, a grey area regarding silent responses remains. Ephratt (2012) has highlighted that silence can have a communicative function demonstrating it may sometimes comply with Grice’s Cooperative Principle once contextual information is taken into account and that silence can potentially trigger implicatures.
Interpretation of silence is not usually a problematic issue in ordinary life, but in a legal context, silence often has serious consequences, some of which are enforced in the Criminal Justice and Public Order Act (1994). The study reported in this paper aims to reassess the usefulness of the legislation particularly the impact of strategical silences based on specific choices rather than the decision to not talk at all.
Using the Harold Shipman trial (2009) as a primary source, a qualitative method is applied to 29 silent responses during the cross-examination. Drawing inferences from silence in a legal setting is challenging, but patterns in the data can be identified. These demonstrate that silence is strategical and has an effective impact on the exchange, particularly in relation to the notion of ‘uncooperative implicatures’.
Key Words: Gricean Pragmatics, Cooperative Principle, Silence, Cross-Examination
An Exploration of the School-to-Prison Pipeline and The Transfer of Behavioural Controls in Adverse Youth-to-Adulthood Transitions.
Lisa Nixon - Keele University
There is an overrepresentation of some demographics in the school exclusion data. This is especially true for boys, some ethnic groups, those with SEND, and those from disadvantaged backgrounds. These overrepresentations also exist within the criminal justice system. This study will explore the School-to-Prison pipeline by examining how the government sanctioned behaviour management techniques employed in education settings may exasperate existing inequalities evidenced in the overrepresentation of populations in the criminal justice system. The methods employed will attempt to address the research on a macro to micro scale. Initially a critical discourse analysis will be employed to track how the object of the policies is observed and analyse the language used. This aims to identify how the wording constructs meaning in the policies, and in turn any underlying social beliefs or inconsistencies in how a child is viewed depending on their position along the pipeline. Secondly the connection to geographical location will be explored, identifying any correlation of statistical data in relation to the overrepresentation of disadvantaged populations in school expulsions and contact with the Youth Justice System. Finally, case studies will highlight the lived experience of people who have experienced behavioural control identified in the policies that follow the School-to-Prison pipeline. A labelling and social control theory theoretical framework will be used to study how the application of the policies may have impacted individual life trajectory.
Quantitative Approaches and Modelling
Transitioning from JHS to SHS in Ghana’s selective entry system: validating students’ meritorious transition belief (MTB)
Nicholas Quartey - The University of Manchester
Ghana implements a long-standing selective school entry procedure for students enrolling into senior high schools. This is based on the quality of the students’ scores in the Basic Education Certificate Examination (BECE). While acknowledging that this process introduces differences in “school-match quality”—defined as the disparity in the quality of school pairings—I posit that this disparity affects students’ belief of what they deserve or do not deserve in terms of school resources and opportunities (i.e., their meritocratic belief: perception of whether they deserve the senior high school they ended in). Consequently, this is hypothesized to influence students’ academic dispositions (i.e., school attachment, achievement motivation, self-efficacy, academic control and resilience) and school outcomes. As part of addressing these dynamics, I will talk about my innovative framework to measure students’ meritocratic belief, rooted in the theoretical imperatives of meritocracy as a desert (with formula S deserves D in virtue of doing A). This framework improves the understanding of meritocratic belief (a cognitive endorsement of meritocracy which remains the traditional measurement approach) by factoring in school attachment (which reflects students’ emotional alignment with the reward of meritocracy) into its measurement to form a proxy for meritorious transition belief. This is a novel construct and has been validated using pilot study data from Ghana and the Rasch measurement, which has shown it to be a good measure. Students’ scores at the measure range from −1.10 logits to +0.89 logits. High scores indicate merit endorsers, and low scores indicate merit skeptics.
Why Does Data Shape Matter?
Simon Rudkin - The University of Manchester
Wherever data is employed there are stories to be extracted and inference to be drawn. Those inferences may drive policy, target interventions or serve as an exploratory data analysis. Statistical models bring functional form to the understanding of relationships, but are premised on the correct specification of the functional form. To know whether models are accurately specified requires further reference to the data. Visualisation is a key component of the understanding of data, of evaluating models and provides a route to target interventions. However, too often empirical work neglects the consideration of the information that exists in the structure of data. Well-understood lessons from Anscombe’s quartet, Simpsons paradox and others are neglected amongst the quest for model fit and explainability. This talk will present new evidence on the importance of thinking about the structure of data. The talk acts as a quick-fire introduction to Topological Data Analysis, a growing area of data science which is founded in the shape of data. The talk looks ahead to how the tools of TDA speak to exploratory data analyses, model evaluation and the development of new modelling tools. Through the 5-minutes we confirm that the shape of data does matter.
Bayesian Conformal Prediction
Fanyi Wu - The University of Manchester
This study examines Bayesian Conformal Prediction (BCP), which integrates Bayesian posterior predictive with non-conformity score in conformal prediction frameworks.
Conformal Prediction (CP) constructs prediction regions with guaranteed coverage properties without distributional assumptions. These regions contain the true outcome with a pre-specified probability (1-α) regardless of the data distribution—a crucial advantage for applications requiring reliable uncertainty quantification. Bayesian prediction provides well-calibrated uncertainty when the assumed model is correct but may exhibit poor coverage under model misspecification (M-open perspective). CP addresses this limitation by providing finite-sample frequentist guarantees without requiring model fidelity.
By integrating these complementary approaches, BCP presents a promising synthesis. Vovk et al. (2005, Theorem2.10) establishes that conformal predictors are at minimum equivalent to conservatively valid predictors. Building upon this property, we investigate how conformity measures derived from Bayesian posterior predictives preserve statistical guarantees while yielding more efficient inference. Caprio, Sale, and Hüllermeier (2023, Proposition 7) further establishes that any conformal prediction region can be refined to generate strictly smaller prediction sets while maintaining identical coverage properties.
These theoretical results suggest pathways for optimising BCP through strategic prior specification and likelihood selection. Our research aims to develop enhanced methodologies that leverage both paradigms' strengths: incorporating structured prior information from Bayesian statistics while preserving the coverage guarantees of conformal prediction. This synthesis offers potential improvements over both standard Bayesian approaches and traditional conformal prediction methods.
What is the Impact of AI Adoption on Firm Productivity? An Empirical Analysis for UK Companies.
Yifan Zhang - The University of Manchester
In recent decades, average technology diffusion rates have increased in all developing and developed countries, and technology adoption has dramatically increased (Diermeier and Goecke, 2017). Artificial Intelligence (AI) is the most prominent, regarded as a new predictive, modular, and general-purpose technology, with a rapid, penetrating, and far-reaching use over many industrial sectors (Brynjolfsson et al., 2021; Aghion, 2018; Czarnitzki et al., 2023). Countries globally also launched many strategies to encourage AI adoption to maintain their competitive advantages because it is expected that adopting AI could increase productivity as an internal driver in various ways (Syverson, 2011). However, techno-pessimists argued that the benefits of AI have been exaggerated because the “Digital economy” is still in the installation phase, infrastructures are not mature yet, and the success largely relies on tangible and intangible complementary assets, which leads to the rise of discussion of modern productivity paradox (Anderton et al., 2023).
This study aims to explore the impact of AI adoption on firm-level productivity, with the focus on the UK region to contribute to this ongoing debate and provide empirical evidence about technology-driven productivity performance. The novelty of our paper resides in analysing the impact of AI on firm productivity by using a novel longitudinal dataset comprising firms involved in AI activities. Specifically, the dataset used in our analysis is constructed by The Data City (TDC), a commercial data creator company. It is expected that the relationship between AI adoption and firm-level productivity is positive.
Systematic Reviews
Qualitative Methods for Eliciting the ‘Voice’ of Preverbal Neurodiverse Young People: A Systematic Literature Review and Qualitative Meta-Aggregation.
Freddie Jones - The University of Manchester
Introduction: The voices of preverbal neurodiverse young people are excluded from research. Reviewing digital methods and assistive technologies which provide innovative approaches for capturing their voices is important, granting researchers and practitioners greater insight over the experiences of non-speaking neurodiverse young people. This may inform research practice developing more inclusive approaches to educational research.
Method: The current study was conducted last year at the University of Manchester and used both systematic and qualitative meta-aggregation techniques examining how non-speaking neurodiverse young people's voices are typically elicited by researchers.
Outcome: The review revealed the dearth of literature which consults the voices of preverbal neurodiverse young people. Two synthesis statements were generated: one related to the Lived Experiences of preverbal neurodiverse young people, and the other reflected how preverbal neurodiverse young people construct their self-concept. Both synthesis statements consisted of 2 categories and 2 sub-categories ranging from transition to autonomy. The synthesis provides practical recommendations for researching preverbal neurodiverse young people, promoting representative research designs targeted at inclusive data collection and analysis. Ultimately The review draws on a coherent and organized framework, using both systematic and synthesis methods to provide a strong foundation to draw “inclusive ” principles for qualitative research with neurodiverse populations demonstrating the value of their inclusion in research.
Implications: Through the application this study generated a bank of research methods and the purposes for which they were use and can now be accessed by fellow researchers to advance research in autistic communication and interaction from a preverbal young persons perspective in order to inform policy change
Conceptualising Adolescent Mental Health within Education Settings: A Systematic Review and Qualitative Synthesis of Cross-Cultural Perspectives
Jiayi Quan - The University of Manchester
Background: Current research on understanding what is adolescent mental health in the context of education research lacks clarity and consensus and should be critically appraised. This systematic review aims to (i) identify and clarify the concepts and terminology related to adolescent mental health commonly used in educational research, (ii) categorise and summarise those definitions under broad conceptual themes, and (iii) develop a cross-cultural perspective on how adolescent mental health is understood in educational contexts in different countries, highlighting both similarities and differences. Method: A systematic review using PRISMA methodology (registered with PROSPERO database) will be conducted. International databases, including British Education Index, ERIC, PsycINFO, PubMed, and Web of Science will be searched, alongside databases widely used in non-English-speaking countries, including CNKI and WanFang databases in China, CiNii database in Japan, as well as KoreaMed Synapse database in South Korea. Publications examining, assessing, explaining, or defining mental health concepts among adolescents aged 10-19 years within educational settings will be included. Methods will involve searching, independent screening, and review using pre-determined inclusion and exclusion criteria, data extraction, coding, iterative content analysis and qualitative synthesis of literature.
Keywords: adolescents; mental health; education; conceptualisation; cross-cultural; systematic review
A New Philosophical Approach to Causal Enquiry: Evidential Pluralism
Jon Williamson - The University of Manchester
Evidential Pluralism is an emerging philosophical account of causal enquiry which holds that we should scrutinise mechanistic studies alongside comparative studies when evaluating a causal claim. This motivates a new approach to establishing causal claims in the biomedical and social sciences, new methods for systematic evidence review, a new approach to mixed methods research, and a new evidence-based approach to the justification of laws. In this talk I give a lightning introduction to Evidential Pluralism and these methodological consequences.
Developing Culturally Sensitive Physical Activity Promotion Strategies for Older Chinese Adults in the UK: An Evidence-, Theory-, and Person-Based Approach
Yang Yang - The University of Manchester
Background: Older Chinese adults are one of the least physically active groups in the UK, but culturally sensitive strategies to promote their physical activity are limited. Behaviour Change Techniques (BCTs) have been used to encourage physical activity (PA) among older adults and could guide the development of PA promotion strategies for this population.
Objective: To identify intervention functions, BCTs and develop culturally sensitive PA promotion strategies for older Chinese adults in the UK.
Methods: The PA promotion strategies were developed following the MRC framework, the Behaviour Change Wheel, and a person-based approach. These strategies were designed based on the findings of a systematic review and interviews with the target population, supported by PPIE meetings.
Results: Six intervention functions (Education, Persuasion, Incentivization, Training, Modelling, Enablement) and 25 BCTs suggested by the BCW were included in the PA promotion strategy package, which was matched with the guiding principles and met the APEASE criteria (i.e., The APEASE criteria, encompassing Acceptability, Practicability, Effectiveness, Affordability, Side-effects/safety, and Equity). A flexible delivery model will accommodate various preferences and capabilities, including booklets, face-to-face workshops, and social media support groups.
Conclusions: Integrating evidence, theory and stakeholders' perspectives enabled the development of a promising culture-sensitive PA promotion strategy package for older Chinese adults in the UK. The findings suggest that future research evaluating these PA promotion strategies is warranted.
Co-authors: Yang Yang, Kimberly Lazo Green, Nan Zhang, Lisa McGarrigle and Chris Todd
Researching Policy Contexts
What Interpretive Social Network Analysis Reveals in Anti-Trafficking Policy
Anna Forringer-Beal - The University of Manchester
This research project proposes an innovative methodology to better understand how policy is implemented. We aim to overlay interpretive analysis of policy documents and interviews with practitioners on top of social network analysis. Using governmental anti-trafficking networks in Manchester and Northern Wales as case studies, we investigate how flows of information across networks are enhanced, derailed, and susceptible to individual discretion on the part of anti-trafficking practitioners. Our aim is to demonstrate how different policy frameworks may shape networks and individual opinion, ultimately resulting in different policy outcomes for people experiencing exploitation. This presentation is a snapshot into our methodology and an examination of how interviews with practitioners may shape the interpretation of a social network.
While this presentation is a single authored piece by Dr Anna Forringer-Beal, this research is conducted in collaboration with the Leverhulme Trust funded research group, ‘Interpreting the Governance of Human Trafficking Law’ (IGHT), which includes Dr Nick Turnbull, Prof Rose Broad, and Prof Tom Vander Becken.
How Can Interdisciplinarity Improve Evidence-Based Methods for Policy Impact Evaluation?
Luana Poliseli - The University of Manchester
A trusted and inclusive science-policy interface is increasingly recognized as necessary for adequate responses to the world's polycrisis. In any context, policies are made by prioritizing certain types of evidence instead of others. For instance, in public health, Orthodox ‘evidence-based’ approaches to policy impact evaluation (EBP) scrutinize experimental and quasi-experimental studies, especially randomized controlled trials (RCTs), excluding other kinds of evidence, e.g. lived experiences. Dealing with this asymmetry, Evidential Pluralism is an emerging philosophical theory of causal inquiry that motivates a new approach to policy evaluation, called ‘EBP+’. According to this method, mechanistic studies should be scrutinized alongside those RCTs that are the focus of EBP evaluations. Although this approach is tailored to the public health context, the inclusive nature and critical reflexity of rethinking evidence assessment make its impact relevant to other policy-making contexts, such as those socioenvironmental related (e.g. climate change, biodiversity conservation, OneHealth, etc.). However, challenges immediately arise when applying the ‘Evidential Pluralism’ method for evidence evaluation in the context of sustainability and socioenvironmental policies. The intractability of most causal relations due to their complexity, combined with the critical social impact of their interventions, makes evidence evaluation sensitive. Thus, how can we use ‘Evidential Pluralism’ for socioenvironmental challenges such as biodiversity conservation and sustainability? More importantly, how can interdisciplinary lessons from environmental sciences and social sciences approaches (e.g. ethnography, embodied methods, etc.) help improve the ‘Evidential Pluralism’ account to become a more inclusive, just, and responsible tool for evidence evaluation in larger complex contexts?
Digital Methods in Climate Policy Research: Unpacking Stakeholder Perspectives in Liberia through Q-Methodology
Vermon Washington - Hertie School
As digitalization progressively reshapes research across the Humanities and Social Sciences, the adoption of computational methods in policy studies presents novel opportunities to address pressing challenges, such as climate change. This research presents a case study from Liberia that explores how digital research tools, specifically Q-methodology, can be utilized to analyze political stakeholder perspectives on climate change and their subsequent translation into national agenda setting.
As climate change impacts intensify globally, Liberia has experienced an annual 20 cm rise in sea levels since the 1980s, with projections indicating a 1.5°C – 2.5°C temperature increase by 2050 (World Bank, 2023). This necessitates urgent, coordinated, and informed policy actions. However, qualitative interviews and discourse analysis alone are insufficient to comprehensively capture the complexity of stakeholder viewpoints in polarized political environments, such as Liberia. Q-methodology, a mixed-methods approach that integrates qualitative statements with quantitative factor analysis, facilitates a systematic mapping of subjective perspectives, thereby providing avenues for analyzing intricate and diverse viewpoints.
By digitizing data collection through online stakeholder surveys and the analysis process through Ken-Q Analysis, this study enhances transparency, reproducibility, and inclusivity. It also addresses ethical concerns pertaining to participation and data ownership. Furthermore, the research reflects on the opportunities and limitations of integrating digital tools in politically sensitive research contexts and offers practical guidance for navigating methodological innovation in the digital era.
This research contributes to broader discussions on digital research ethics, stakeholders’ collaborative production, and the methodological challenges associated with working at the intersection of sustainability and public policy.
The Application of AI in Academic Writing
Minjie Xing - The University of Manchester
This empirical study explores the application of AI in students’ essay writing, focusing on its impact on essay quality, student engagement, and students’ perceptions and suggestions regarding this AI-assisted writing mode. The research employed a mixed-methods approach, collecting quantitative data through pre-and post-test assessments and qualitative insights via student reflections. Thirty-five advanced-level students in a Chinese-degree program participated in the study. Thematic analysis of the qualitative data identified key themes related to student reflections perceptions, including benefits, challenges, and recommendations for more effective AI integration. Quantitative findings revealed a significant improvement in essay scores, particularly among students who actively engaged with AI, compared to those with minimal interaction. Students reported that AI was helpful in generating ideas, providing relevant information, suggesting appropriate vocabulary and idiomatic expressions, and fostering critical thinking skills through targeted prompts. However, they also noted limitations such as information overload, mechanical or generic responses, occasional inaccuracies, and a lack of cultural nuance. Based on their experiences, students recommended a more structured and pedagogically guided use of AI. Key suggestions included the need for precise prompts, critical engagement with AI-generated content, and a clearly defined role for AI in the writing process. They also advocated for AI’s integration into the curriculum to support the development of more creative and imaginative writing skills. This research contributes to the growing field of AI-assisted language learning by offering practical recommendations for effectively integrating AI into teaching, learning, and research.
Understanding NFT through the Walkthrough Method: Applying Digital Sociology Method to Understand Fintech Platform
Ernestina Xinyi Zhu - The University of Manchester
With the advancement of technology in cryptocurrency, non-fungible tokens (NFT) have become more and more accessible to users. Any user could register an account in a cryptocurrency wallet and start trading NFT instantly. This research is intended to use the Walkthrough method to understand how users interact with one another on the NFT trading platform. The Walkthrough method (Light, Burgess and Duguay, 2018) is intrinsically a qualitative method in digital sociology to investigate how users interact with one another on digital platforms. Instead of focusing on the technical infrastructure solely, it emphasises human-machine interactions and how social interactions could play a role in economic behaviours. As for this research, I will go through each interface of OpenSea (the biggest NFT trading platform), including the registration profiles, logging-in process, users’ daily usage, and logging-out process. Each function and symbol will be analysed through the users’ perspectives and documented through screenshots and digital diaries. In that case, the affordances of the website and how people make decisions on the NFT trading platform could be observed. As for future research, I argue that it requires more research in customer behaviour studies to follow up and verify the observation of NFT trading platforms.
References:
- Light, B., Burgess, J. and Duguay, S. (2018) ‘The walkthrough method: An approach to the study of apps’, New Media & Society, 20(3), pp. 881–900. Available at: https://doi.org/10.1177/1461444816675438.