Marking the first time the premier international conference has taken place in Europe after previous iterations at Harvard, Duke, Michigan, and other universities, the event brought together leading researchers and practitioners from across North America, the UK, Europe, and beyond.

A Vibrant Programme of Methods and Research

The four-day conference showcased cutting-edge methodological developments and empirical applications in political network analysis:

• 114 participants attended across the four days.

• 96 research presentations were delivered across 30 sessions in 4 parallel panel tracks, spanning topics from policy networks and political polarization to international relations and online discourse.

• 8 intensive training workshops took place over the first two days, covering introductory and advanced network analysis in R, causality-aware temporal networks, deep graph learning, data collection and management, and statistical modelling.

• Featured sessions included a mentoring lunch, workshops with senior academics, and a publishing panel with the editors of Network Science, where a dedicated special issue for the conference is hosted, and Nature Human Behaviour.

Strong Local Engagement and Institutional Partnerships

The event saw substantial local participation, with 25 attendees from the University of Manchester taking part thanks to conference stipends provided by the Cathie Marsh Institute for Social Research (CMI). North American graduate students were supported by 10 APSA travel grants sponsored by the APSA Political Networks Section.

The conference was organised in partnership with the Department of Social Statistics, the Mitchell Centre for Social Network Analysis, the Cathie Marsh Institute for Social Research, the APSA Section on Political Networks, the ECPR Standing Group on Political Networks, and the International Network for Social Network Analysis (INSNA). Generous support was provided by the University of Manchester's Hallsworth Conference Fund, the Mitchell Centre, and the CMI.

For full access to session abstracts and the conference archive, visit the POLNET 2026 website.

Abstract

Political actors form coalitions around their joint normative beliefs in order to influence the policy process on contentious issues such as climate change or population aging. Policy process theory maintains that learning within and across coalitions is a central predictor of policy change but has yet to explain how policy learning works. This article explains the formation and maintenance of coalitions by focusing on the ways actors adopt policy beliefs from other actors in policy debates. A policy debate is a complex social system in which temporal network dependence guides how actors contribute ideological statements. Belief adoption matters in three complementary ways: bonding, which exploits cues within coalitions; bridging, which explores new beliefs outside one's perimeter; and repulsion, which reinforces polarization between coalitions and cements their belief systems. We formalize this theory of endogenous coalition formation in policy debates and test it on a micro-level empirical data set.

Read the article in the American Journal of Political Science: https://onlinelibrary.wiley.com/doi/10.1111/ajps.70079

Abstract

Behaviour by individuals or organizations is often interdependent. Social contagion posits that behaviour spreads from unit to unit due to the presence of network or equivalence relations as transmission pathways. Contagion of a single behaviour has been modelled in cross-sectional and temporal data contexts. But existing statistical approaches have not been able to identify multiple contagion pathways in temporal processes where multiple actors can display or adopt multiple behaviours. This data structure and problem setting is common, for example in health behaviours by peers, treaty ratification by states, the spread of wildfire incidents in forests, or the diffusion of policies or political beliefs.

We explore the application of bipartite relational event models of actors and behaviours and find that temporally backward-looking specifications confound social contagion with prior similarity, the tendency of similar units to adopt the same behaviour independently. We construct a set of sufficient statistics parsing information bidirectionally along the event sequence to establish an atemporal prior similarity null distribution against which contagion hypotheses for multiple pathways can be tested. Using simulations and four empirical cases, we show the efficacy of this parametric approach for disentangling contagion from prior similarity, contributing to causal inference for temporal networks.

Read the article in the Journal of the Royal Statistical Society, Series A (Statistics in Society): https://academic.oup.com/jrsssa/advance-article/doi/10.1093/jrsssa/qnag069/8699341.

Abstract

Relational event models (REMs) can infer the generative properties of longitudinally observed social networks with instantaneous edges. They assume conditional independence of edges given sufficient network statistics formed over the past event sequence. A popular specification in REMs is to subject these statistics to exponential temporal decay with a fixed half-life parameter to attribute higher importance to more recent edge events in the formation of network statistics. Assuming a fixed half-life parameter may cause biased estimates and obfuscates the temporal horizon over which network effects operate in empirical social systems. These limitations are addressed by proposing fully Bayesian estimation of REMs and designating the half-life parameter as an estimable quantity. A “pre-computation” strategy is devised to speed up calculations for practical feasibility of the sampling procedure.

The approach is adapted to discourse network analysis, which models political actors’ statements about their preferred policy beliefs as dynamic networks. An application to the policy debate on reforming the German public pension system illustrates how temporal decay for inertia, actor activity, belief popularity, and actor homophily can be estimated alongside the main coefficients. Convergence diagnostics and an illustration of bias correction relative to fixed parameters are provided, stabilisation using hyper-parameters and computational complexity are discussed, and the approach is extended to include both Breslow’s and Efron’s methods for breaking ties in the event sequence.

Read the article in Computational Statistics and Data Analysis. https://www.sciencedirect.com/science/article/pii/S0167947326000976?via%3Dihub

Abstract

The European Union faces long-term governance challenges in contested domains, such as migration management, health data sharing, and facial recognition technology. Across these fields, political debates are shaped by shifting ways in which actors frame problems and solutions. Understanding how such framing contests evolve over time is crucial for explaining both integration dynamics and the capacity of EU institutions to govern effectively.

This article develops the mathematical foundations of a comparative approach to frame evolution by extending discourse network analysis, acknowledging the relational nature of collective action framing. Partitioning trajectories into distinct states and phases reveals key types of frame evolution: progressive versus regressive, entangled versus disentangled, and fast-paced versus slow. This typology provides a basis for linking framing dynamics to EU integration and policymaking processes. The article makes three contributions: operationalising collective framing trajectories; introducing a temporal kernel smoothing approach; and identifying comparative dimensions across policy debates.

Read the article in the  Journal of Common Market Studies: https://onlinelibrary.wiley.com/doi/10.1111/jcms.70119

A central aim of the initiative has been to make data and digital skills feel more accessible and less intimidating for students across different disciplines, particularly for students who may not initially see themselves as “technical”. Rather than focusing only on coding or software, the project emphasises collaboration, transparency, peer learning, and confidence-building.

Dr Tatjana Kecojević, Lecturer in Social Statistics and Director of the Manchester Q-Step Data Fellowship Programme, said:

Recent activities have included the Data4All Hangout, Introduction to R workshops, reproducibility sessions, GitHub portfolio workshops, and collaborative mentoring activities where students work together on transparent and reproducible workflows.

Students involved in the initiative contribute not only as participants, but also as organisers, GitHub contributors, peer mentors, workshop facilitators, and collaborators. The project therefore places strong emphasis on co-creation and student leadership.

The initiative was also recently presented at the School of Social Sciences Scholarship Showcase, where five students co-presented alongside staff, each showcasing their own activities, GitHub portfolios, and reflections on their experiences with mentoring, reproducibility, collaboration, and digital skills development.

The presentations were exceptionally well received, with colleagues across the School praising the students’ confidence, professionalism, collaborative work, and the quality of their portfolios and contributions.

Several students also reflected on how these activities helped strengthen their confidence, presentation skills, sense of community, and employability.

Dr Kecojević added:

Alongside its teaching and learning focus, the initiative is also beginning to develop wider civic and international connections.

One of the next goals is the development of a Q-Step Civic Data Lab initiative that would connect students, community partners, policy organisations, and applied data practice through inclusive and scalable learning pathways.

The wider relevance of this work is also being recognised internationally through the proposal “Inclusive Data Practice Lab, developed collaboratively with partners from the Office for IT and eGovernment in Serbia, which is selected for the UN World Data Forum 2026 programme.

Dr Kecojević said:

The initiative also highlights the growing importance of reproducible and transparent workflows in both academic and professional contexts. By encouraging students to document and reflect on their analytical processes through GitHub and e-portfolios, the project helps students develop transferable skills increasingly valued by employers.

Future plans include expanding mentoring opportunities, developing Civic Data Lab collaborations, strengthening civic and policy partnerships, and creating further opportunities for students to engage with applied and socially impactful data practice.

Further information:

• Data4All: https://dataforall.netlify.app/

• Data4All 2025: https://data4all2025.netlify.app/

• SoSS GitHub Organisation: https://github.com/SoSS-UoM

• Reproducibility resources: https://rrmcrsoss.netlify.app/

• Introduction to R resources: https://sossintroductionr.netlify.app/

• Scholarship Showcase slides: https://canva.link/7adgd7acudw41k7

Papers included:

• Understanding Masculinity: How it is Defined, Measured and Experienced.
• Growing Up Without Parents: Spatio-Temporal Dynamics of Skipped-generation Households and the Impact on Adolescent Mental Health.
• The Division of Labour and Female Partners' Relative Pay Across Phases of Parenthood: Evidence from UK Couples.
• The Decomposition of Group Differentials.
• Working Life Expectancy Among China’s Middle-Aged and Older Population.
• How Does the Use of Information and Communication Technologies Impact on the Wellbeing of Older Adults?
• Got SWAG? A Simple Scalable Approach to Bayesian Inference.
• How Effective is the UK Homeless Policy in Assisting Young People Out of Homelessness?
• Exploring the Impact of Gender Roles and Social Norms on Fertility Intention Among Young Adults - An Empirical Analysis of International Data.
• A Socio-Semantic Analysis of the Perception of Large Language Models by Scientists.
• Factor Recovery of Bifactor Models Using Machine Learning.
• Work Schedule Typologies and Job Satisfaction Among UK Workers.
• Willingness to Participate in Smartphone-Based Research: Evidence from the UK.

Find out more about our current PhD students and their research projects.

On Wednesday, 29 October, the Social Statistics Department in SoSS hosted the Data4All event, bringing together students, alumni, academics and industry professionals to explore the power of data skills in education, research and employment.

The morning session, led by Dr Tajana Kecojevic, saw a full room of students from the University of Manchester and Manchester Metropolitan University eager to learn about reproducible research skills and how to create an e-portfolio. Due to the success of the session, repeat workshops are being organised to meet the student demand to learn more about showcasing their skills for the future.

The afternoon session shone a spotlight on the 380 students who have been Q-Step Data Fellows since the programme launched in 2013. Set up and funded by the UK government, Q-Step is designed to address the shortage of quantitative data skills taught across university humanities courses, particularly in the social sciences.

The Manchester Q-Step Centre is one of 15 centres across the country, and each year places second-year undergraduate students in organisations ranging from local charities in Manchester to government departments in London. The data fellows are paid the living wage to complete an eight-week data-driven, research-led work placement, designed to focus mainly on quantitative skills. Since the programme started, at least 70% of students placed each year have been female, with 25% having come from underrepresented backgrounds or disadvantaged groups.

At the Data4All event, students who were placed in data fellowships this summer came to present their work through posters and presentations. The former Q-Step Programme national lead, Dr Simon Gallagher, spoke about the impact the programme has made on the UK higher education sector and the benefits to students, employers and universities in developing data skills through humanities and social science degrees.

Two of our Q-Step lecturers reflected on the importance of experiential learning to the development of the curriculum and past data fellows shared with us their career trajectories, highlighting how Q-Step helped them get there. Employers who hosted students were full of praise for their talent, creativity and the passion they have for applying their classroom learning in the workplace. A workshop led by Unit-M led to ideas for the future sustainability of the programme and reflected the demand for data skills in the workplace.

It was an especially emotional afternoon for Professor Jackie Carter, who is leaving the Data Fellows programme this year after pioneering it in 2013. Having placed over 380 students in around 70 organisations over the years, Jackie is still in touch with many alumni.

She reflected on the success of the programme: “In 2013 I had an idea. That was to open doors to data careers to students studying social science and humanities degrees through paid work placements. The Data Fellows programme was born. It has grown into an award-winning initiative, and in 2025 my baby is about to become a teenager.

“Having been at the start of some of our graduates' careers has been one of the highlights of my own career at the University. But what I am most proud of is that it is a programme that supports social mobility and inclusion. As I pass the baton on to my successor, I hope they will honour the legacy. It’s been an honour and a privilege to create a lasting impact.”

Professor Fiona Devine, Vice-President and Dean of the Faculty of Humanities, has championed the scheme since the start, helping to initiate it at the University. Fiona said: “Like Jackie, I am deeply humbled and very proud that the Q-step initiative has been so successful. Quantitative skills are very important for all social scientists and I am delighted that we have facilitated the acquisition of these highly sought after skills.

“The work placement opportunities have been excellent, and I am so grateful to all those partners who helped us. We have a wonderful Q-Step alumni community pursuing great careers in a wide range of sectors of the economy.

“A huge thanks to Jackie for her outstanding leadership. Her passion and commitment have driven this success.”

The project will advance current statistical methods for survey inference, enabling more profound exploration of entropy-based diagnostics, improved modelling strategies, and the development of accessible tools for analysts working with complex multi-source datasets. The award also facilitates collaboration with colleagues across institutions and will lead to a series of academic outputs and practitioner-focused guidance materials.

Overall, this grant represents a significant opportunity to advance methodological innovation in the field and support more reliable, evidence-based decision-making in the social sciences. 

Find a list of British Academy/Leverhulme Small Research Grants Awards 2025 here.

Abstract

In private capital investment, limited partners (LPs) and general partners (GPs) frequently encounter the challenge of finding suitable counterparts amid limited information, a process often hindered by market inefficiencies. This article addresses this issue by exploring the micro-level mechanisms that shape private capital networks, employing temporal exponential random graph models. Our findings uncover activity and popularity effects, persistence mechanisms, and homophily in preferences concerning region, strategy, and industry. These factors jointly shape the dynamically evolving network structure across asset classes and the hybrid network with all asset classes, revealing a shared network formation process. This article offers practical insights into the matching problem within the private capital market.

Read the article in Social Network Analysis and Mining

Flow matching is a generative modelling technique that learns to transform random noise into meaningful data by following smooth trajectories. Think of it as a more flexible and efficient cousin of diffusion models. This approach has become a backbone for many modern generative models across different domains - from image generation to, in this case, synthetic tabular data.

The collaborative environment at AMLab fostered innovative discussions that refined the methodological framework and strengthened the international research network between the Department of Social Statistics and the University of Amsterdam. Working at one of Europe's leading machine learning research labs provided a unique environment to engage with cutting-edge computational and quantitative research, especially in synthetic data generation.

The findings from this work are now available as a preprint on ArXiv, representing a significant milestone in advancing generative models for tabular data, especially to provide privacy-preserving and high-quality synthetic data. This collaboration exemplifies the Department's commitment to fostering global academic partnerships and contributing methodological advancements to the broader machine learning and statistical community.

Isabella's master's thesis from Queen’s University in Canada focused on how peer support levels and problematic social media use of youth in military families differed from non-military-connected youth. This project used the largest Canadian dataset with responses from military-connected youth themselves, the HBSC survey. Youth in military families have unique experiences that can shape their health outcomes, which is why this research is important. 

Isabella was awarded the Colonel Russell Mann Military, Veteran and Public Safety Families Research Award for my research “Online communication and problematic social media use among military-connected youth in Canada”. This award recognises high-quality Canadian research that deepens the understanding of the health, resilience and lived experiences of military, Veteran and public safety families.

The focus of the visit was to contribute to the development of a novel methodological framework that integrates Matrix Decomposition-based (MD) estimation into Structural Equation Model (SEM) Trees and Forests.

Traditional SEM Trees rely on Maximum Likelihood Estimation (MLE), which can be unstable, have improper solutions (e.g., negative variances), and be computationally intensive, especially in small subgroups or misspecified models. The team in Tokyo worked on developing an alternative approach based on matrix decomposition, which avoids many of the pitfalls of likelihood-based estimation.

During his visit, Ahmed worked on extending existing simulation studies from single-tree models to ensemble-level forest models, comparing four major estimation frameworks: MD-based SEMTree, Maximum Likelihood-based, constrained ML-based, and Bayesian SEMTree. He implemented forest-level simulations that assess both shared metrics (such as improper solution rates, computational time, and node recovery) and forest-specific ones (like variable importance, prediction accuracy, and ensemble diversity). He also continues to explore alternative splitting algorithms, including Factor Analysis by Instrumental Variables (FABIN) and other non-iterative multi-start approaches. Eventually, the team intends to develop an open-source R packages to support this new methodology.

This visit provided a unique environment to engage with cutting-edge computational and quantitative research which contributes to the methodological advancements that will benefit the broader SEM and statistical community.

Ahmed will continue to collaborate with Professor Usami and Dr Todo not only to finalize and publish this research, but to collaborate for more research together in the future. The Department of Social Statistics at the University of Manchester and the team at University of Tokyo both expressed interest in more contact and collaboration in the future between the departments and the universities in general.

Examining how young people across Europe imagine and plan their path to adulthood, the article, “Ideations and Intentions in the Transition to Adulthood: A Cross-European Comparison,” draws on data from the Generations and Gender Survey and the European Social Survey to explore young adults’ ideal ages, intentions, and actual behaviours across 33 European countries. 

The study provides a comparative picture of when young people expect to leave home, form partnerships, marry, and become parents; and how these expectations align, or fail to align, with reality.

Findings show that while young Europeans tend to view their twenties as the ideal period for key life transitions, they often experience these events later than intended. The mismatch between ideals and outcomes highlights persistent gender and regional differences, suggesting that cultural norms and structural barriers continue to shape the pathways to adulthood across Europe.

Primary health services are the first point of contact for those seeking support for common mental health problems, such as anxiety and depression. These services form an essential part of the response to the UK’s mental health crisis. 

Engagement with those accessing and navigating primary care pathways for common mental health problems has highlighted that accessing and engaging with support frequently involves struggling with complex and burdensome bureaucratic processes at an already difficult time. These experiences can leave people with a profound sense of unjust treatment that continues throughout their journey through the system.

Research across different disciplines has suggested that perceptions of ‘just’ treatment can also significantly impact treatment outcomes by affecting engagement with services, deterring individuals from seeking support in the future, and exacerbating existing inequalities within the system.

A research programme, led by Professor Joe Tomlinson and funded by a £2.5 million Wellcome Discovery Award, will pioneer the application of the theory of ‘bureaucratic justice’ in primary mental health services. By deploying an innovative combination of a longitudinal quantitative and qualitative study with participatory storytelling methods, the programme will develop a new understanding of the nature and relevance of just treatment in the context of how people interact with frontline administrative processes in primary care mental health services.

The programme will also build an interdisciplinary community of researchers, while providing a platform to embed lived experiences of seeking mental health support into research practices in innovative and engaging ways.

The programme formally launches in January 2026 and will run for six years. Alongside Professor Joe Tomlinson, The Dickson Poon School of Law and King’s College London, the programme will benefit from the expertise of co-investigator Professor Todd Hartman Head of the Department of Social Statistics and Deputy Director of the Cathie Marsh Institute at The University of Manchester. Co-investigators are also based at the University of York (including Dr Jed Meers, Dr Simona Manni, Dr Annie Irvine, Dr Aisling Ryan, and Professor Lina Gega).

The research, funded by Good Neighbors (in 2022-23) and The University of Manchester - International Science Partnerships Fund (in 2024-25), was led by a joint team from the School of Social Sciences and Global Development Institute, University of Manchester, including Dr Jihye Kim, Professor Wendy Olsen, Dr Mohammed Ibrahim, Harshada Ambekar, Sonny McCann and Mindy Park. 

They conducted both surveys and focus groups, finding that students who participated in the programme showed a shift in their attitude about early pregnancy, such as postponing the expected age of having their first child. The programme’s success is attributed to its emphasis on education, future development, and mutual respect among peers, rather than solely focusing on negative outcomes.

Community leaders have observed a noticeable drop in teenage pregnancy rates. This progress is encouraging, but the report emphasises the need for continued support and a long-term strategy. Misunderstandings about family planning still need to be addressed at a community level to achieve a sustainable reduction in teenage pregnancy.

The study concludes that social engagement is a potent tool for empowering young people and improving sexual and reproductive health knowledge.

The baseline and follow-up study reports are available on the Good Neighbors UK website. The paper on the results from the baseline study has been released in Children & Society.

We gratefully acknowledge the support of the Good Neighbors Alliance, known globally as Good Neighbors International, a federation of independent but affiliated Non-Government Organisations operating in over 50 countries.

You can read the paper, “Up and Down With… Polarisation? Intrinsic and Instrumental Polarisation Dynamics in US Climate Policy Debates”, online.

Political elites in the US are ideologically divided over climate change. We identify two perspectives: 

• The intrinsic view on ideological climate polarisation views climate beliefs as entrenched parts of an actor’s identity and posits that ideological positions have factually shifted towards ever more extreme positions over time.
• The instrumental view, in contrast, emphasises that polarisation entrepreneurs mobilise their constituency to participate in the climate policy debate by amplifying ideological differences over climate-related focusing events when they arise, leading to fluctuations in visible polarisation, rather than a steady trend. 

This study examines which of the two perspectives holds in US Congressional and subnational media debates by analysing time trends of polarisation and phases of structural stability. We distinguish between endogenous events, which can be attributed to the political process, and exogenous focusing events, such as extreme events or those related to the international climate regime, and investigate which type of event tends to be associated with changes in polarisation. 

Applying two novel time series measures for discourse networks - structural polarisation and the detection of phases of structural stability - to the climate debate during the 112th to 114th Congress (2013–2017) and subnational print media in four swing states, we find that exogenous events are largely irrelevant while endogenous political dynamics increase the polarisation of the debate considerably. 

We find ups and downs of polarisation corresponding to distinct structural phases in which polarisation is linked to participation. This temporal fluctuation of polarisation around endogenous political events is consistent with the instrumental perspective.

Drawing on their expertise in population data, large scale social surveys, and the use of data to inform fair and effective health policy, the Manchester academics highlight the vital role of robust evidence in ensuring policy keeps pace with societal change.

Launched in 1991, the Health Survey for England, has been the backbone of health policy evidence for more than three decades, providing annual, nationally representative data to monitor the nation’s health and guide healthcare delivery. At the June 2025 conference, the government confirmed that NHS England would no longer run the survey. While details of any replacement are yet to be confirmed, discussions are under way on the future of population health surveys in England.

Since the conference, The UK Data Service has acted swiftly to ensure researchers’ voices are heard in the national debate:

• Boosting consultation responses – negotiating with Department of Health and Social Care (DHSC) and the Office for National Statistics (ONS) to re-open a user consultation on the future of population health surveys for three weeks in July, generating a further 140 responses from health survey users.
• Showcasing impact – collaborating with a key health survey user to publish a Data Impact blog highlighting the importance of the survey: Help shape the future of population health surveys.
• Supporting decision-makers – at DHSC/ONS’s request, preparing a written summary of the Health Survey for England’s value and impact, including user testimonies, to feed into their internal report (due end of September), and advising on an expert peer-reviewers.

Through this work, Manchester academics and the UK Data Service are playing a pivotal role in shaping the future of the Health Survey for England, influencing how health data will be collected to meet the challenges of a changing world.

Practical sessions included the implementation of Cox Proportional Hazards models and Kaplan-Meier estimators using R. A highlight was a constructive meeting with Professor Yabiku and colleagues concerning future work and collaboration on research on migration.

KOSTAT-UNFPA Summer Seminar on Population

The Summer Seminar on Population was first launched by the East-West Center (EWC) in 1970 and quickly gained recognition as a leading population seminar series. In 2013, responsibility for population activities was transferred to Statistics Korea (KOSTAT), which hosted the first KOSTAT Summer Seminar on Population in 2014.

Since 2017, the seminar has been co-hosted annually by KOSTAT and the United Nations Population Fund (UNFPA), continuing its mission to advance population studies and statistical capacity building.

The seminar serves as an international platform for government statisticians, graduate students, and population experts to exchange ideas, share research, and strengthen statistical capacity in population-related fields. Over the years, participants from more than 20 countries have taken part in the programme.

Read the paper, “Single and partnered mothers’ labour market consequences of long family leave”.

This study examines the heterogeneous labour market effects of family leave policies for single and partnered mothers. 

Longer family leave has been shown to weaken women’s labour market positions and some studies have found heterogeneous effects across population groups. However, whether the effect differs by partnership status remains unexplored. 

Using Finnish register data from 1989 to 2014 (ca. 2.5 million person-years) and controlling for selection into single motherhood by comparing estimates from OLS and FE models, this study compares single and partnered mothers’ unemployment and earnings consequent to extended family leaves. In line with predictions that single mothers may face greater work-family reconciliation issues or cumulative disadvantage leading to greater labour market penalties, the results showed that longer leave increases the length of unemployment for single mothers more than for partnered ones. 

This is not solely because of selection into single motherhood. Earnings penalties after family leave (net of employment status) are the same for single and partnered mothers. 

We conclude that similar long- lengths of family leave are penalised more among single mothers in terms of employment, which increases and reproduces social inequalities. This means that existing inequalities are reinforced by labour market absences supported by leave policies.

The ESRA Outstanding Service Award acknowledges sustained and high-level contributions to European survey research, either of a methodological, substantive or infrastructural nature.  Nominations are made and voted on by members of the ESRA Committee.

Natalie publishes widely in areas of survey statistics and survey methodology, including survey design and estimation, adaptive survey designs, small area estimation, non-probability sampling, data linkage and integration, confidentiality and privacy. 

She is an elected member of the International Statistical Institute (ISI), a fellow of the Royal Statistical Society, a fellow of the Academy of Social Sciences and President 2023-2025 of the International Association of Survey Statisticians.  She also serves on editorial boards and international Methodology Advisory Boards at National Statistical Institutes.

In her acceptance speech for the award, Natalie noted that survey methodology and survey statistics are becoming increasingly important. She emphasised the need for high-quality randomised probability-based survey data to evaluate accuracy and mitigate biases in non-survey data sources, such as administrative data, big data and non-probability samples. She mentioned that only through the knowledge and understanding of the theoretical underpinnings of statistical methods and inference can we truly move forward into the digital and AI era.

Policy subsystems are comprised of competing advocacy coalitions, in which public and private political actors with shared belief systems learn from each other and coordinate their strategies in the pursuit of influencing policy making in their favour. 

While numerous studies have focused on the longevity and structural stability of advocacy coalitions, there is scant theory and evidence on how nascent policy subsystems bifurcate into stable, competing coalitions. 

This article proposes a three-stage model of problem discovery, differentiation, and consolidation. 

We apply discourse network analysis to the nascent subsystem of the UK's COVID-19 response in order to study these phases and discuss their applicability and implications for other institutional and issue contexts.

Our colleagues, Eduardo Fé & Mario Pezzino, have published a study in Decisions in Economics and Finance. Read the paper: “The long term implications of co-creation in economics education”.

Co-creation - where students help design teaching materials - has clear short-term benefits for engagement and soft skill development. But our new research shows it also creates powerful intertemporal peer effects: students exposed to co-created materials become more motivated, feel part of a learning community, and are more likely to co-create themselves.

We develop a dynamic model of how co-created resources influence student effort over time and test this through a behavioral experiment in an intermediate microeconomics course. The results suggest that co-creation not only deepens learning, but can gradually reshape education culture—boosting what we call "education morale."

Findings

• Co-creation promotes responsibility, reciprocity, and intrinsic motivation
• Exposure to peer-created materials increases future engagement
• Even small interventions (e.g., emails from former students) can create measurable, lasting effects.

The paper “Goodbye human annotators? Content analysis of social policy debates using ChatGPT” can be read online.

Content analysis is a valuable tool for analysing policy discourse, but annotation by humans is costly and time consuming. ChatGPT is a potentially valuable tool to partially automate content analysis for policy debates, largely replacing human annotators. 

We evaluate ChatGPT’s ability to classify documents using pre-defined argument descriptions, comparing its performance with human annotators for two policy debates: the Universal Basic Income debate on Dutch Twitter (2014–2016) and the pension reforms debate in German newspapers (1993–2001). We use the API (GPT-4 Turbo) and user interface version (GPT-4) and evaluate multiple performance metrics (accuracy, precision and recall). 

ChatGPT is highly reliable and accurate in classifying pre-defined arguments across datasets. However, precision and recall are much lower, and vary strongly between arguments. These results hold for both datasets, despite differences in language and media type. Moreover, the cut-off method proposed in this paper may aid researchers in navigating the trade-off between detection and noise. 

Overall, we do not (yet) recommend a blind application of ChatGPT to classify arguments in policy debates. Those interested in adopting this tool should manually validate bot classifications before using them in further analyses. At least for now, human annotators are here to stay.

The paper “When does discursive change happen? Detecting phase transitions in discourse networks of sustainability transitions” can be found on ScienceDirect.

Sustainability Transitions Research (STR) confronts complex societal challenges by examining societal shifts and their trajectories. An emerging perspective in STR is discursive approaches, which analyse the role of discourses and discourse coalitions in shaping sustainability transitions. However, discursive approaches face challenges regarding the analysis of sustainability transition processes as complex, temporal processes of stability and change. 

We discuss the nature of these challenges and extend the method of discourse network analysis (DNA) by measuring distinct temporal states (phases of stability) in discourse networks and detecting phase transitions (significant changes) between these discursive states. 

Whereas most approaches analyse discursive changes in a top-down way, we introduce a method for the bottom-up detection of discursive stability and change. This facilitates a more accurate tracing of how sustainability transitions unfold over time. An empirical application of this extension to the discursive networks around the introduction of a Low Emission Zone demonstrates how and when discourses and actors display significant structural shifts. This methodological innovation addresses the need for measuring stability and change in the complex, discursive, temporal dynamics of sustainability transitions.

Highlights

• Identification of 3 challenges of discursive approaches to energy transitions.
• Discourse Network Analysis is extended by drawing on mathematical network science.
• Bottom-up detection of temporal phases, states, transitions in discourse networks.
• Energy transitions should be analysed through the lens of temporal networks.
• Case study on the changing debate on a low emission zone in Ghent, Belgium.

You can read the paper, “Multiregional Population Forecasting: A Unifying Probabilistic Approach for Modelling the Components of Change”, on Springer.

Regional population forecasts are important for planning and understanding how populations are changing and redistributing. To forecast regional population changes, one must have a mechanism to capture different sources of population growth. In low fertility and developed societies, the main factors driving population redistribution are internal migration and immigration, for which both tend to concentrate people towards large metropolitan areas. 

In this article, we extend the multiregional cohort-component population projection model developed by Andrei Rogers and colleagues in the 1960s and 1970s to be fully probabilistic, by using Bayesian inference. We apply the model to forecast population for eight states and territories in Australia. 

The projections are based on forecasts of age-, sex- and region-specific fertility, mortality, interregional migration, immigration and emigration. The approach is unified by forecasting each demographic component of change by using a combination of log-linear models with bilinear terms. 

This research contributes to the literature by providing a flexible statistical modelling framework capable of incorporating the high dimensionality of the demographic components over time.

Forecasts of a population totals by sex for states and territories in Australia, b total population. States or territories: NSW—New South Wales, VIC—Victoria, QLD—Queensland, SA—South Australia, WA—Western Australia, TAS—Tasmania, NT—Northern Territory, ACT—Australian Capital Territory. 

Prior to starting the position, Mariana was working on her PhD in Sociology at the University of Oxford. Her PhD was titled “A Sociodemographic Study of Only-children and Only-child Fertility” and explores only-child fertility from a sociodemographic perspective. 

At the University of Oxford she also worked as a research assistant on the project “Digital Gender Gaps”, focusing on combining traditional surveys and big data for population analysis. Her main research interests are family demography, fertility and the use of formal demographic methods for exploring changes in kinship networks.

Since joining The University of Manchester, she has been co-teaching in SOST10012-Understanding Social Media Data and SOST2002-Essentials of Survey Design and Analysis.

Prior to moving to the UK, Mariana trained as a demographer at the Centre for Development and Regional Planning (Cedeplar) in Brazil, where she was awarded an MSc, and developed her research interests in family demography, as well as formal demographic methods.

The award is presented annually to students who demonstrate exceptional contributions to artificial intelligence (AI) and data science research.

Constanza's doctoral research employs advanced machine learning techniques to examine the causal impact of front-of-package food labelling on consumer behaviour, drawing on both randomised trials and observational data. Her research in this field has implications in areas ranging from health causal inference to data linkage.

She will receive funding to continue the research and join the Turing Institute network, a collaborative platform aimed at advancing AI and data science.

Congratulations to Constanza on this remarkable achievement!

Further information on the Turing Enrichment Scheme can be found on The Alan Turing Institute website.

“Ground-truthing political elites in the public sphere: Measuring the arena effects of elite opinion” is available to read online.

Political elites express their ideological positions on contentious issues across various arenas in the public sphere. Social science research often relies on data extracted from various media or political and administrative sources, as well as surveys that are administered directly with the political actor. 

Although some studies compare ideology across different sources, few systematically analyse how political actors adjust their ideological messaging to the audiences in the respective communication arenas and how such changes are associated with systematic bias in data sources. 

This paper uses a unique dataset, combining climate policy belief observations from three arenas - social media, Congressional testimony, and surveys - on identical ideological variables and during the same time period. 

We apply item response theory to understand how responses differ by arena and find that ideological communication on X is most left-leaning, Congressional testimony is most right-leaning, and surveys, the data source with the smallest potential arena effect, is in the middle. We also find that actors with strong ideological leaning moderate their positions on social media and in Congress. 

These findings enhance our understanding of strategic communication depending on audience context and inform social research on biases when analyzing specific data sources.

I am a Professor of Social Statistics in the Department of Social Statistics at The University of Manchester, which is part of the School of Social Sciences. I am also a member of the Mitchell Centre for Social Network Analysis and the Cathie Marsh Institute for Social Research. 

Before joining Manchester in April 2024, I was a Professor of Comparative Politics in the Department of Government at the University of Essex. 

I am also a DFG Mercator Fellow in the Research Training Group on Digital Platform Ecosystems at the University of Passau (2022-2027) and serve as chair of the Political Networks Section of the American Political Science Association (APSA) in 2024-2025. 

My main research interests are politics and public policy, network analysis and complex systems, statistical modelling, and computational social sciences. I am best known for my work on discourse network analysis (and the software Discourse Network Analyzer), the texreg R package, and my work on statistical models for longitudinal network data (e.g., implemented in the btergm R package). 

My research has appeared in leading political science journals (e.g., AJPS, JOP, BJPS), public policy and administration journals (e.g., PSJ, JPART), technical journals (e.g., Physica A: Statistical Mechanics, Network Science, JStatSoft), and outlets in other fields (e.g., Nature Climate Change, Addiction, Personality and Social Psychology Bulletin). 

More information on Professor Philip Leifeld can be found on his website. 

The paper, “Learnings from a decade of data fellows: Co-creation of a data skills framework for non-stem students”, is available to read online. 

This workshop provided a reflection on an experiential learning model developed in the UK. The Data Fellows initiative supports undergraduate social science and humanities students to develop their data skills through work placements. 

The findings have resulted in a book, academic articles and international presentations which collectively provide a substantial body of evidence to illustrate how non-STEM (science, technology, engineering and maths) students can learn and practice their data analytic skills and progress into data and technical careers. 25% of the 373 Data Fellows placed to date have been from historically under-represented groups and 70% have been female. 

A case study was presented to show a journey from a first degree in social science to a postgraduate degree in data science. The aim of the workshop was to challenge the deficit narrative that can accompany the teaching of data skills in the social sciences and explore whether a suitable data skills framework exists or could be developed.

For more information about Jackie visit her research profile.

Slides and recordings (if available) from these events can be accessed on the event web page by clicking the links above. Recordings are also available on the UKDS YouTube channel. 

For information about upcoming events visit the UK Data Service website.
 

Our colleague, Dr Nikita Basov, has recently published a study, “The Space of Coordination: Accounting for Multiple Expert Knowledges in Environmental Communication”, in Environmental Communication. 

The authors propose an account of environmental communication as a dynamic space involving multiple expert knowledges to address growing diversity of expert knowledges in environmental communication. 

To enable this account, they offer a computer-assisted mapping technique relating these knowledges to each other at various time points. 

The authors illustrate the proposed approach with a case study on flood risk management in the UK, where diverse expert groups have been engaged in a shared communication space which enabled coordination of their knowledges over time. 

They conclude that researchers can use the proposed technique to trace knowledge dynamics in environmental communication. Communication practitioners can use it to map thematic areas that experts specialise in, identify knowledge gaps, find relevant documents, and facilitate expert communication. 

The project “Building computational capacity among global data service staff” was funded for the UKRI's call on “Digital infrastructure: new approaches to skills or software."

Digital research infrastructures connect researchers, policymakers and innovators with the computers, data, tools techniques and skills to undertake ambitious and creative research. 

Social science data services worldwide play a key role in the digital research infrastructure by curating and managing access to many forms of social and economic data as well as promoting increased data literacy among the community. 

Recognising the growing importance of computational skills for data services staff in the social sciences, this project, led by academics affiliated with the UKDS, will address the critical need for training. 

The aim of this project is to build capacity within the international data services community, by providing upskilling opportunities for UK Data Service (UKDS) staff and developing foundational level data skills modules in computational social science for the wider global community. It will also establish a community of practice to provide enhanced support to users through the lifetime of the project and beyond. 

Direct beneficiaries of this project will include UKDS staff who will be given the opportunity to upskill in computational skills, as well as global data service staff who will be given access to a foundational-level online structured course(s) on computational social sciences. 

Through both upskilling mechanisms, this project will enhance data services capacity both in the UK and globally, enhance the careers of data service professionals, and through the establishment of a Community of Practice will contribute to a culture of lifelong learning.

Attendees explored how sociotechnical changes might shape the future of research, hearing from a variety of experts and collaborating with researchers and professionals from different sectors and disciplines. 

In addition to fostering vibrant discussions, Methods Con: Futures 2024 featured the launch of NCRM’s Futures Briefs a series of publications reflecting on how sociotechnical changes might impact the way social science is conducted. 

The event was a resounding success, garnering excellent feedback for its interdisciplinary approach, valuable insights and providing an opportunity to engage with cutting-edge methods. 

The NCRM is proud to have facilitated such a dynamic exchange of ideas and looks forward to its continuing commitment to staying at the forefront of methodological innovation in the social sciences. 

During the month of September, University of Manchester colleagues at the UK Data Service (UKDS) delivered the following events: 

• Computational social science drop-in (10 September, J. Kasmire, Louise Capener and Nadia Kennar) 
• Data in the spotlight: Large scale social surveys (24 September, Nigel Denoronha and Placide Abasabanye) 
• Survey data drop-in (24 September, Alle Bloom, Pierre Walthery) 
• Presentation at RSS International Conference 2024: Coding in public: recognising vulnerability as a barrier to reproducibility (2 September, J. Kasmire) 
• Presentation at RSS International Conference 2024: Developing quantitative data driven research skills in social science students outside the classroom (2 September, Vanessa Higgins) 
• Presentation at NCRM methods showcase: The UK Data Service Data Skills Framework (11 September, Vanessa Higgins)

Slides and recordings (if available) from these events can be accessed on the event web page by clicking the links above. Recordings are also available on the UKDS YouTube channel.

For information about upcoming events visit the UK Data Service website. 

The Department of Social Statistics celebrates the publication of the report “Exploring the Intersection of Technology and Democracy”. 

Led by Professor Mark Elliot, the 2023 Futures Summer Camp, funded by SPRITE+, was a collaborative effort with the Research Institute for Sociotechnical Cyber Security (RISCS) and facilitated by the School of International Futures (SOIF). 

The event aimed to foster future-focused collaborations across various disciplines to address challenges in digital security, identity, privacy, and trust.

The camp brought together 28 experts from academia, government, and industry for a two-day workshop to explore how technological advancements might impact democratic institutions over the next 5-15 years. 

Participants discussed potential threats and opportunities, considering actions to mitigate negative outcomes and maximize positive impacts.

Key themes and insights from the workshop will shape SPRITE+’s future initiatives, including a TIPs-focused expert meeting in May 2024 and a themed sandpit in June 2024 titled "Living in an Inauthentic World". 

This event underscores the importance of interdisciplinary collaboration in preparing for the future intersection of technology and democracy.

The report can be accessed here.

On Thursday 16 May, the department of Social Statistics held its away day at HOME, in central Manchester. The event was a resounding success, fostering a collegial environment where important discussions flourished. 

Key themes included research collaborations within the department and support for early career academics. The day’s activities not only sparked meaningful conversations but also played a crucial role in shaping the future direction of the department.

During the month of April, University of Manchester colleagues at the UK Data Service (UKDS) delivered the following events:

• Computational Social Science Drop-in (9 April, J. Kasmire, Nadia Kennar and Louise Capener)
• How to become a computational social scientist (16 April, Louise Capener and Nadia Kennar)
• N8 CIR Synthetic Data Generation Event (17 April, J. Kasmire)
• SRMS Dissertation workshop: Locating Research Data and Data Access (18 April, Vanessa Higgins)
• Getting Started with Secondary Analysis (22 April, Nigel de Noronha)
• DRTN Lessons from Covid-19 data (30 April, J. Kasmire and Alle Bloom)

Slides and recordings (if available) from these events can be accessed on the event web page by clicking the links above. Recordings are also available on the UKDS YouTube channel.

For information about upcoming events visit the UK Data Service website.

Professor Mark Elliot (Social Statistics) has been awarded an ESRC grant for the project Enhancing Data Accessibility and Security through Innovative Data Synthesis (EDASIDA).

The EDASIDA project aims to transform both data accessibility and confidentiality through innovative data synthesis techniques. In essence, the project will develop a methodology for providing tailored teaching datasets and systematic disclosure risk assessment methods.

The new methodology involves leveraging cleared analytical outputs from data services as the basis for generating synthetic data using genetic algorithms. The goal is to provide trainees with data that not only closely resembles real-world data but also yields analytical output very similar to that of the real data, enhancing the training experience. 

A pilot study conducted in collaboration with Administrative Data Research UK, demonstrated the feasibility of generating synthetic teaching datasets with both high utility and no marginal disclosure risk. The pilot dataset (a synthetic version of the linked ASHE-census dataset) was successfully used in a ADR-UK training course in April 2024.

The approach also offers a route to formalise assessment the disclosure risk associated with analytical outputs from safe settings. By embodying statistical outputs in synthetic data, it enables a systematic evaluation of disclosure risk, addressing the informality and potential inconsistencies present in current output checking procedures. 

Finally, the project aims to bolster the federated services agenda by exploring the creation of synthetic linked data from using analytical outputs from data of multiple services. This approach expands the possibilities of data synthesis without the need for actual linkage and elaborate governance of infrastructure, such as trusted third parties.

On Wednesday 24 April, the Department of Social Statistics hosted the grand finale of the Data Science in Practice Series. 

Karolina Michalska from KANTAR's London office captivated an audience of eager students from the MSc Social Research Methods and Statistics program and the undergraduate Data Analytics pathways.

Karolina dazzled the attendees with an exhilarating presentation showcasing her groundbreaking work at KANTAR. 

She offered an in-depth and eye-opening overview of the diverse and dynamic roles available for data scientists within the company. Her presentation was not only highly informative but also sparked a vibrant and engaging discussion, leaving everyone inspired and buzzing with ideas.

Professor Philip Leifeld (Department of Social Statistics) has published a new article in the prestigious journal Social Networks. 

In the paper, Leifeld and his colleagues present a Bayesian framework for testing scientific expectations in Exponential Random Graph Models. 

Their framework intends to overcome some of the limitations affecting classical settings, such as inconsistent behaviour when the null hypothesis is true, their inability to quantify evidence in favour of a null hypothesis, and their inability to test multiple hypotheses with competing equality and/or order constraints on the parameters of interest in a direct manner. 

To tackle these shortcomings, this new publication presents Bayes factors and posterior probabilities for testing scientific expectations under a Bayesian framework. The methodology is implemented in the R package BFpack. The applicability of the methodology is illustrated using empirical collaboration networks and policy networks.