Past projects

Termeh Shafie, David Schoch, Pete Jones, Elisa Bellotti, Martin Everett

The project aims to make a valuable contribution to the analysis and modelling of multivariate networks in which network data comprises multiple relations of different kinds, together with various node attributes measured on different scales. This research consists of an intertwined theoretical and practical part where the former is aimed to be both descriptive and inferential.

In short, the research includes (i) using multigraph representations to allow a simultaneous analysis of composition and structure, (ii) statistical entropy analysis to extract observed and latent interdependencies, both within and between composition and structure. We are currently focusing on finding novel interdisciplinary application areas and extending the framework to ego, signed and two-mode networks.

Pete Jones, Termeh Shafie

This one-year funded ESRC Postdoctoral Fellowship explores the use of quantitative network methods for analysing gender inequalities in popular film texts. The tools and perspectives of social network analysis offer an interesting lens for representing fictional narratives as relational data consisting of interactions between characters. Building on doctoral research on the same subject, the project aims to explore the extent to which narrative dynamics can be captured through character network dynamics. In particular, it will focus on developing a relational event model approach for asking questions about the gendered patterns of interaction between characters in popular cinema, such as whether female characters are less likely to interact with each other than they are with male characters.

Elisa Bellotti

The project focuses on domestic violence and the role of the personal network of abused women in the decision to call the police. As well as the interpretation of the consequences of this decision. When the police are called, various parties in the network have on occasions tried to hinder and prevent violence. The starting point of the proposed investigation is to study the ‘response networks’, that the persons in the troubled relationship are embedded in, with the focus on the meaning and social consequences of involving the police.

Nick Crossley

Over ten years, working both individually and in various collaborations, Nick Crossley has been exploring the networks involved in music making. Focusing upon punk, post-punk, folk, heavy metal and Two Tone, and taking in networks of gigs and festivals as well as musicians and audiences, he has explored the structure of such networks; their formation and evolution; and their sociological importance in allowing music to 'happen'. The two main outputs of this project are:

• Crossley, N. (2015) Networks of Sound, Style and Subversion: the punk and post-punk worlds of Manchester, Liverpool, London and Sheffield, 1975-1980 (Manchester University Press).
• Crossley, N. (2020) Connecting Sounds: the social life of music (Manchester University Press).

Tomás Diviák and Nick Lord

Deferred Prosecution Agreements (DPAs) are a legal tool for the non-trial resolution of cases of bribery and corruption. In the UK, the Serious Fraud Office has the power to negotiate DPAs with corporations implicated in bribery. Each DPA is accompanied by a Statement of Facts that provides detailed and publicly available textual records of the given cases, including summarized evidence of who was involved, what they committed, and with whom. These statements can be translated into networks amenable to social network analysis allowing an analysis of the structure and dynamics of each case.

We extract information about (i) which actors were involved in a given case, (ii) the relations and interactions among these actors (e.g., communication or payments), (iii) their relevant individual attributes (e.g., affiliation or position), (iv) the timestamp of the interactions, and (v) the institutional context of each case from each Statements of Fact. We code the extracted information manually with two independent coders and subsequently, we assess the inter-coder reliability. Using SNA, we construct nine bribery networks, describing the structure and resilience of each network, its evolution over time, and test the effect of various mechanisms underlying its structure and dynamics.

ESRC grant ES/R004471/1 (£315,613.61)

PI: Rosemary Board (University of Manchester). Co-Is: David Gadd (University of Manchester), Elisa Bellotti (University of Manchester), Carly Lightowlers, University of Liverpool

The problem of human trafficking has received increased attention in the last 20 years, but it was only in 2015 that the UK adopted a systematic legislative framework, namely the Modern Slavery Act 2015 (HM Government, 2015). The act has tackled the rising cases of human traffickers and managed to increase the convictions, also widening the focus of policing and prosecution which now includes not only sexual exploitation but also servitude, labour exploitation and forcing people to engage in fraudulent and criminal activities, such as coerced begging and false benefit claiming (Dwyer et al, 2011).

Academic research has however been sparse, and little is known about traffickers’ motives, networks and backgrounds. Our project “The Perpetrators of Modern Slavery Offences: Motivations, Networks and Backgrounds” funded by the Economic & Social Research Council (ES/R004471/1, 2018-2021), aims to fill this gap by bringing together the analysis of different type of data to provide an initial understanding of the networks, trajectories and personal experiences of convicted human traffickers in the UK.

We analyse data on victims, suspects and crimes recorded by Greater Manchester Police (GMP) during the period from April 2015 to June 2018 in its Modern Slavery Data Tool (MSDT), plus a separate dataset relating to referrals to GMP from the National Referral Mechanism (NRM) for the same time period. The NRM is the formal process through which victims are identified as victims of modern slavery in England and Wales. Descriptive statistical analysis is accompanied with social network analysis to explore the size and composition of criminal groups convicted for modern slavery offences.

The quantitative data are complemented by 14 qualitative interviews conducted with people convicted for modern slavery. Interviews provide empirical materials to dig into the trajectories and experiences of traffickers and show a more nuanced picture of the criminal motives and opportunities. Gendered labour and constraining social relationships indicate how offences are conducted and motivated within the social circles of families, colleagues and neighbourhood which offer criminal opportunities and often chain down offenders in increasingly unavoidable illicit paths.

National Institute of Health (US) grant ($33,004)

Programme director: Jane Carlton, NYU. Lead of project: Sandra Albert, Martin Luther Christian University; Director, Indian Institute of Public Health – Shillong. Co-I: Elisa Bellotti, University of Manchester

The goal of this project is to conduct mixed methods, social network-based, socio-behavioral studies to determine what types of malaria-preventive interventions reduce malaria through good coverage and use, and to identify demand and/or supply side barriers to malaria control measures in the northeastern state of Meghalaya, India. In this area the malaria situation is dynamic. Cases steadily increased from 2012-2015, but since then there has been a downward trend, and 2017 had the fewest cases ever reported. This steep decline has been attributed to the distribution of long-lasting insecticide impregnated bednets (LLINs) in 2016 and partial adoption of indoor residual spraying (IRS) with DDT. Nevertheless, residual cases remain endemic, and to design an appropriate elimination strategy, we must try to understand the reasons for these cases. Data from the Meghalaya Department of Health Malaria Division and from the Center for the Study of Complex Malaria in India (CSCMi) ongoing epidemiology studies suggest that community resistance to IRS, misuse of LLINs, unprotected outdoor activity, and hidden reservoirs of asymptomatic, submicroscopic, and hypnozoite infections all contribute to the endemic persistence of malaria. At Barato PHC (primary health center) and Nonglang PHC in Meghalaya, 10 villages in each site will be subjects of epidemiological studies funded through the parent CSCMi 2.0 grant. In Aim 1 we plan to examine how malaria-preventive measures are adopted in practice, using ethnographic methods of non-participant observations; in Aim 2 we will explore the cultural and social barriers preventive measures may face, using qualitative interviews and focus group discussions (FGDs); and in Aim 3 we will assess which community members are most likely to influence villagers to accept or reject such measures, using social network analysis surveys. When these Aims have been concluded, their findings will be summarized and integrated with those of other CSCMi studies (epidemiology, vector) underway at the Barato and Nonglang PHCs into a malaria situation analysis for the villages involved. This will yield a better understanding of the focal malaria epidemiology, as well as barriers and opportunities to current interventions used (e.g., LLINs, IRS), and possibilities for additional interventions (e.g., treatment of household members of malaria cases). It will also enable us to provide recommendations to the state malaria control program for additional measures that are likely to foster or improve the effectiveness of IRS and LLINs, with the long-term objective of complete elimination of malaria in Meghalaya.

ESRC IAA grant £24,552

PI: Nick Lord, University of Manchester). Co-I: Elisa Bellotti (University of Manchester)

ScriptNet is a software package that facilitates an analysis of the organisational aspects of criminal enterprise together with an analysis of the network of people, organisations, places and resources that are in some way involved in the commissioning of these goal-oriented crimes. The tool can be used by a variety of groups - including law enforcement authorities, non-governmental organisations, academic and social researchers, and many more - to visualise the connections between the different stages of pre-planned criminal behaviours and the people or organisations who play different roles, in different places, using different resources to accomplish specified criminal goals. ScriptNet is an amalgamation of the terms ‘script’ and ‘network’ that in turn represent two analytical approaches to understanding criminal and social behaviours: crime script analysis and social network analysis. ScriptNet integrates these two analytical techniques to provide a structured and systematic approach for researchers and investigators of goal-oriented crimes. By using ScriptNet, you will be able to draw upon your knowledge of known and on-going criminal cases to visualise the entire network of people, organisations, places and resources that are involved at key stages during the commissioning of the crime. You will also be able to visualise which people play which roles during the criminal enterprise, to visualise the personal networks of individual people and their relationships to others in the criminal enterprise and visualise the geographical aspects of the criminal enterprise. These different layers of analysis will in turn inform your investigative processes by reinforcing your insights and enabling you to visualise your thought processes, or by generating new lines of inquiry into particular people or places.

The software is a collaboration between the University of Manchester (Nicholas Lord, Elisa Bellotti and Cecilia Flores Elizondo), Joshua Melville and Steve McKellar (Team Garlic). The ScriptNet software is available for use with Windows and Mac. Download the release here. Inside the ZIP file is an installer for the software. Double click it, and the software will install and then open. The software is not digitally signed, which might mean that Windows/Mac tells you that it is unsafe. You should force it to install the software by telling Windows/Mac you want to install anyway. Once it is installed, you can open the software by looking for “ScriptNet” in your start menu or in your Mac applications list.

Cite as follows: Lord, N., Bellotti, E., Flores Elizondo, C., Melville, J. and McKellar, S. (2020) ScriptNet: An Integrated Criminological-Network Analysis Tool, University of Manchester.

Elisa Bellotti

There is a prevalent narrative that the corona virus pandemic is a great leveller and ‘we are all in this together’. Indeed it does seem that the virus does not respect wealth or privilege, as Hollywood stars (Tom Hanks), political leaders (Boris Johnson), royalty (Prince Charles) and world champions (Lewis Hamilton) succumb to its effects. However, it would be simplistic to assume that the pandemic and its associated restrictions impact on everyone in the same way.

There is mounting evidence of the salient influences of gender, class, age and ethnicity in how the pandemic is experienced across society and how it is acting as a multiplier of inequalities. While there has been much research on health outcomes, there is need for more analysis of the social impact of current restrictions.

Therefore, during the summer of 2020, as lockdown was beginning to ease, a team of social scientists from the Universities of Manchester, Nottingham, East Anglia and London Metropolitan University launched a nationwide survey to investigate whether and to what extent personal relationships had changed during the lockdown, and how confinement had impacted upon domestic, work, leisure and social habits.

The research team includes: Dr Elisa Bellotti, Mitchell Centre for Social Network Analysis, University of Manchester. Dr Alessio D'Angelo, International Centre for Public and Social Policy (icPSP), University of Nottingham. Professor Louise Ryan, FAcSS, Director of the Global Diversities and Inequalities Research Centre, London Metropolitan University. Dr Emilie Vrain, Department of Environmental Science, University of East Anglia. The team was supported by Professor Dan Rigby, School of Social Sciences, University of Manchester, who set-up the online questionnaire using the Lighthouse Studio Sawtooth Software platform.

The preliminary report (February 2012) is available on the Lockdown Networks website.

This was funded by the University of Melbourne.

András Vörös, Termeh Shafie, David Schoch, Martin Everett, Johan Koskinen, Yoshihisa Kashima, Garry Robins, Hugh Gallagher.

Social influence is crucial in explaining individual behaviour at all levels of education. Social rewards and punishments from peers, friends, and parents shape students’ school attitudes, substance use, and other lifestyle choices. Advances in statistical network modelling has allowed researchers to analyse social influence in the context of changing social relations at school. However, traditionally, data and methods are limited to small-scale settings, up to a few hundred students. This ignores the wider context of influence represented by families, neighbourhoods, and local communities. We aim to extend state-of-the-art dynamic network models to be applicable to large-scale data with several thousand students, datasets increasingly collected recently. We explore potential extensions based on two suitable educational datasets from a Swiss university and a Swedish town. We design a large-scale empirical network study on school attitudes in the UK and Australia and prepare a grant proposal to fund this research in the future.

Reducing social contacts in schools to attenuate the spread of COVID-19: Which strategies are effective and what are their unintended consequences?

Elisa Bellotti, Tomáš Diviák, Martin Everett, András Vörös

Following the recent wave of the COVID-19 epidemic, reopening schools in the UK are implementing social contact reduction strategies to protect the health of their students, teachers, and families. These measures will influence the school experience of students over the next academic year and, possibly, beyond. Despite their importance for health and learning, three key aspects of contact reduction strategies are unexplored: 1) their relative efficiency in tackling the spread of the virus, 2) their sensitivity to rule-breaking behaviour and 3) their differential impact on the lives of students from various social backgrounds. To address these issues, we propose to examine the efficiency of four social network-based approaches to contact reduction in schools and identify the students who may be adversely affected by them. Combining network simulation models with existing social network data from 214 schools in England, we analyse the potential spread of COVID-19 in real-world settings under different contact reduction strategies. Our findings will help us engage with UK school leaders and educational policy-makers. We will assist them in designing contact reduction strategies that are effective in slowing the disease and fair in distributing their social costs equally among students and families.

Publications

• Bellotti, E., Vörös, A., Diviak, T., & Everett, M. (2020). Reducing social contacts in schools to attenuate the spread of Covid-19. Which strategies are effective? Working paper accepted and published by the Education Committee of the UK Parliament, paper no. CIE0406. URL: https://committees.parliament.uk/writtenevidence/9119/html/
• Written evidence submitted by - committees.parliament.uk CIE0406. Written evidence submitted by Dr. Elisa Bellotti, Dr. Tomáš Diviák, Prof. Martin Everett and Dr. András Vörös Reducing social contacts in schools to attenuate the spread of COVID-19. Which strategies are effective? Elisa Bellotti, András Vörös, Tomáš Diviák, Martin Everett committees.parliament.uk (retrieved: 08/09/20)

András Vörös

The importance of social relations for various individual outcomes in education is increasingly recognised. However, significant gaps remain in our knowledge about which social ties affect student outcomes and how. To date, network studies of student groups tend to ignore that social relations in school are (a) dynamic and change on multiple time scales, (b) multidimensional (multiplex), (c) embedded in the wider context of families and local neighbourhoods.

To overcome these limitations of current research and better understand the role of social networks in education, we aim to:

• collect multidimensional and longitudinal network data in UK schools using novel survey tools;
• extend existing dynamic network models to large networks consisting of heterogeneous actors (students, teachers, parents, etc.) and ties (school friend, parent, sibling, playground friend, neighbour, etc.);
• use these methods and data to understand the impact of different social ties on students’ behaviours, school attitudes, and well-being;
• work together with schools and local policy-makers to develop new educational practices that can improve the school experience and achievement of all students.

Publications:

• Vörös, A., Boda, Z., Elmer, T., Hoffman, M., Mepham, K., Raabe, I. J., & Stadtfeld, C. (2021). The Swiss StudentLife Study: Investigating the emergence of an undergraduate community through dynamic, multidimensional social network data. Social Networks, 65, 71-84.

Boda, Z., Elmer, T., Vörös, A., & Stadtfeld, C. (2020). Short-term and long-term effects of a social network intervention on friendships among university students. Scientific Reports, 10(1), 1-12.

• Stadtfeld, C., Vörös, A., Elmer, T., Boda, Z., & Raabe, I. J. (2019). Integration in emerging social networks explains academic failure and success. Proceedings of the National Academy of Sciences, 116(3), 792-797.

Martin Everett, Nick Crossley, Gemma Edwards, Johan Koskinen, Susan O’Shea, Gemma Edwards, Chiara Broccatelli, James Coutinho

Our aim is to collate an archive of covert social networks and test a number of theories about the organisation of covert networks in addition to exploring a number of methodological aspects particular to the study of illicit networks (find out more on the project, funded by the Leverhulme Trust). In particular, we focused on the dynamics of covert actors and their affiliations to events and organisations, and how we can leverage co-participation and co-membership to learn incomplete data. A PhD on this project (Broccatelli) also conceived of a novel analytical approach for tracing co-participation over time named a Bi-Dynamic line-graph.

Johan Koskinen

This was a small grant (funded by BA/Leverhulme Trust) aimed at facilitating collaborative work with the Universities of Melbourne and Swinburne. Incorporating dependencies stemming from context, locations, and memberships into the modelling of partially observed networks. We developed a suite of approaches for using Bayesian data-augmentation to account for different types of missing data.

Johan Koskinen, James Coutinho

We are working with Canadian Law Enforcement to use statistical network analysis techniques (including multilevel ERGM) to understand the behaviour of organised criminals and criminal organisations operating across Canada. Using data supplied by Canadian Law Enforcement, we are also developing methods for fitting ERGMs for very large networks.

Johan Koskinen

The Multilevel network modelling group (funded by the Leverhulme Trust under the International Collaborative Networks scheme) was a multi-national, multi-institution consortium for developing a new network paradigm for networks with multiple types of ties on multiple types of nodes. We contributed to defining the paradigm and the organisation of a number of meetings and symposia. Our collective conclusions are reflected in a recent Springer book (edited by Emanuelle Lazega and Tom Snijders) outlining the various forms of multilevel networks, best methods, and challenges. Named on or with formal research agreement: Network structure and social processes in empirical social systems (ARC); Complex Ties: The Role of Networks in the Commercialisation of Public Research (ARC); Social Influence of Dynamic Networks (ECRP VI).

Alcohol Research UK, £58,907

Elisa Bellotti

The purpose of this project is to provide a greater understanding of the distribution and the market in counterfeit alcohol. This project investigates the social networks that are engaged in the distribution of counterfeit alcohol. This distribution activity requires a high level of organisation and a developed network of actors to ensure market penetration. Understanding whether interdependency between each of the distribution points exists is a key element in understanding the effective penetration of the market with counterfeit alcohol. Understanding the process of distribution will also provide a means of understanding the localised nature of how counterfeit alcohol distribution is organised and how these locally organised groups are networked with the wider networks of distribution.

A pilot project - The University of Manchester, £24,903

Elisa Bellotti

This pilot project integrates a criminological and social network analytical theoretical approach to understand the organisation of the distribution of counterfeit alcohols. We analyse the dynamics between the ‘scripts’ through which offenders must go in order to accomplish their counterfeit alcohol enterprise and how these scripts are shaped by the networks of cooperating actors at various stages of the crime commission process.

Publications:

• Bellotti E., Spencer J., Lord N., Benson K., Covert and overt networks in Counterfeit Alcohol Distribution: a Criminological Network Analysis, British Journal of Criminology, forthcoming.
• Lord N., Spencer J., Bellotti E., Benson K., A Script Analysis of the Distribution of Counterfeit Alcohol Across Two European Jurisdictions, Trends in Organized Crime, forthcoming.

Elisa Bellotti

The project aims to analyse the scientific collaborations to PRIN projects in Italian Academia from 2001 to 2010. We focus on the factors that increment the chances of getting funded, the interdisciplinarity of a research project and the gender differences in the structural position of collaborations.

Publications

• Bellotti E., Kronegger L., Guadalupi L., 2016, The evolution of research collaboration within and across disciplines in Italian Academia. Scientometrics, 109, 2, pp. 783–811.
• Bellotti E., Guadalupi L., Conaldi G., 2015, Comparing fields of sciences: the network of collaborations to research projects in Italian academia, in Lazega E., Snijders T., (eds.) Multilevel and network analyses, Methods Series, Springer, pp. 213-244.
• Bellotti E., 2012, Getting funded. A multi-level network of Physicists in Italy, in Social Networks, 34, pp. 215-229.

Gemma Edwards in collaboration with the Morgan Centre for Research into Everyday lives

‘Under the Same Roof’ (ESRC), with Morgan Centre colleagues about contemporary forms of shared living. Gemma looked in particular at the role of personal networks in shared living contexts, like housing cooperatives and cohousing schemes, and used time diary data from participants to analyse the social networks of sharers.

Pilot project: Do patterns of friendship formation predict student satisfaction and academic achievement in higher education?

Daniel Tischer and David Hughes

Our research project builds on previous studies investigating friendship formation in educational settings but seeks to provide novel contributions. The study combines the psychometric measurement of participant personality with longitudinal network data. Personality is a driver of friendship formation and success and collecting this data allow us to examine whether or not different personality traits are associated with different network patterns. Spatial data regarding student position within the classroom (eg front, back, left, right) will be analysed to see if certain types of personality lead to different seating positions (do the attentive students sit at the front and the ‘skivers’ at the back?) and also whether the choice of seating is related to other network behaviour.

Daniel Tischer and Adam Leaver

The project advances an alternative explanation to the financial crisis from a social network perspective. In contrast to existing explanations of bubbles focusing on demand-side and investor-behavioural accounts of bubbles, we are seeking to portrait developments in the supply-side, or in other words, how do financial markets grow to become what is widely recognised as a “bubble”? We are using the US market for CDO structuration as a case study to give a time-map of actors entering (and exiting) the market and to show that actors’ positions within the market may inhibit market efficiency where the presence of apparently risk-mitigating actors is nullified by the fact that they sourced and managed asset generated by a highly concentrated core.

Daniel Tischer, John Hassard, Paula Hyde and Leo McCann

As part of this large research project I am organising the collection and analysis of cross-sectional ego-network data on middle managers in different firms and industries, and across six countries. The aim is to better understand how work-related social networks impact on middle managers, both in a supporting capacity (advice, trust, status) and in a stress-inducing capacity (negative ties).