Archives

A personal view on the Netflix hit, Adolescence

A Warning for Parents, a Teaching Moment, or Just a Drama?

by City St George’s UoL doctoral researcher Sylwia Wypyska-Kieran

I finally got around to watching the show. I braced myself for it. I work in this field, and when I sit down to watch TV, I want an escape. And more importantly, I have a son the same age as the main character. I was scared—I know how dramas can pull the right emotional strings to shake you to your core. And this show does exactly that.

It grips the audience, tapping into their deepest emotions to climb the charts. It spreads fear about youth violence, fueling the anxieties of parents who are already overwhelmed by the mysteries of the online world. At the school gates, friends tell me how upsetting it was. Online, acquaintances share tips on how to ‘better control’ our children, while colleagues publish their expert opinions. Following the discourse surrounding the show is fascinating. But let’s start from the beginning.

Katie was murdered. In a typical narrative about crime, the show’s focus is on the perpetrator – a 13-year-old boy. This compelling drama has done an amazing job of avoiding the othering of the boy who killed, a common and widely criticised practice in the media. Jamie looks young and innocent. His child-looking face shatters the stereotype of a perpetrator, which is so important for society’s understanding of the reality of everyday violence. We feel for him. We see a child whose life has crumbled. We feel for his parents, we feel their pain, self-blame and disbelief. Together with the detectives, we seek the answers. Why?

I was expecting the answer to be the manosphere. The trailer and discourse surrounding the show heavily focused on that. The online world of incels and Andrew Tates. I was surprised and rather confused to see that the manosphere was not a direct effect on Jamie’s behaviour. Katie was bullying Jamie, calling him an incel and telling him that no one will ever go out with him.

How did a drama about a boy who murdered a girl manage to make her seem responsible for his crime? Whilst I pondered whether I misunderstood the implication, I saw a comment online ‘What the boy did was definitely wrong but didn’t it start from the bullying by the girl!!!’.

The detective’s son pulls his Dad aside and tells him about the secret language of teenagers. I work with some amazing people whose work is all about young people’s participation. Teenagers don’t live on The Planet of the Adolescents, to which we have no access. Yes, adolescence is a distinctive period in people’s lives and we do have to understand the psychosocial challenges young people face. But let’s not align adolescence with violence and let’s not separate ourselves from them. It is harmful to them, to us, and to society as a whole.

‘Adolescence made free for schools as Keir Starmer meets creators’, the BBC reports, alongside calls to introduce anti-misogyny lessons. But we can’t teach our way out of misogyny. It seems like a reactive decision made without consultation with experts or young people. The context, complexities and consequences of this decision could be immense.

The boys I have spoken to, as part of my research exploring responses to harmful sexual behaviours, have told me that they feel shut down in conversations about relationships and sex. It is consistent with other studies around the topic. They already feel they do not have a space where they can explore and learn about relationships or their identities as men. Will showing this series to them open up a conversation or prove them right? My bet is on the latter. We risk pushing them further into a corner—driving them toward the very spaces where they do feel heard.

References

Youngs, I. (2025) ‘Adolescence hard to watch as a dad, Starmer tells creators ‘, BBC News, 31 March. Available at: https://www.bbc.co.uk/news/articles/cx28neprdppo (Accessed: 02 April 2025).

Gooch, B., Cooke, M. (2025) ‘Schools to run anti-misogyny classes for boys in bid to tackle toxic masculinity’, The Independent, 25 March. Available at: https://www.independent.co.uk/news/uk/home-news/school-misogyny-classes-boys-toxic-masculinity-adolescence-b2718706.html (Accessed: 04 April 2025)

King-Hill, S (2025) ‘Adolescence in schools: TV show’s portrayal of one boyhood may do more harm than good when used as a teaching tool’, The Conversation, 02 April. Available at: https://theconversation.com/adolescence-in-schools-tv-shows-portrayal-of-one-boyhood-may-do-more-harm-than-good-when-used-as-a-teaching-tool-253158 (Accessed: 02 April 2025).

For further information, please contact Sylwia at sylwia.wypyska-kieran@citystgeorges.ac.uk

Photograph from Adobe Photo Stock subscription

Researching the impact of Black and Asian women leadership within East of England domestic abuse services

The East of England is a region with minimal presence of ‘by and for’ (BFR) domestic abuse (DA) specialist services despite being home to Black and Asian communities. A VISION-funded research project, ‘Nothing about us without us’: Investigating the impact of the leadership of ethnic minority women on domestic abuse service provision in East England’, is exploring the impact of the leadership of Black and Asian women within DA service provision in the region.

As part of the work, researchers Dr Mirna Guha (Anglia Ruskin University) and Dr Katherine Allen (University of Suffolk), hosted a leadership event on 3 April 2025 for racially and culturally diverse women. Held at a venue provided by the City of London police, the event was part of a leadership programme implemented through the HUM (‘Us) : A Place-based Emerging-Leaders Model designed and piloted by Mirna and Katherine to diversify leadership in domestic abuse and sexual violence services. Research aims include ensuring culturally responsive and representative support for minoritised victims-survivors in East England.

Prior to the April event, Mirna and Katherine researched the leadership needs of 19 overstretched frontline practitioners i.e. ‘emerging leaders’ from racialised communities working within White-majority and at times professionally isolating generalist services across Bedfordshire, Suffolk and Cambridgeshire. East England is home to scattered racialised and diasporic communities where women’s leadership in local politics and policymaking is low. Based on this, three events on trauma-informed, culturally responsive and diverse styles of leadership have been co-produced with national DASV experts. Participants were invited to also join a regionally pioneering Community of Practice (CoP).

At the April event in London, 15 emerging leaders from racialised backgrounds gained insights into the strategies, opportunities and challenges of practising culturally and racially representative leadership through an inspiring keynote by Superintendent Jasvinder Kaur, Domestic Abuse Lead at Suffolk Constabulary and co-founder of the National Women of Colour in Policing network. Furthermore, nine Black and racialised women speakers associated with statutory institutions viz. Cambridge City Council and the National Police Chief’s Council.

Voluntary organisations also spoke including Apna Haq in Rotherham, Asian Women’s Resource Centre in London, and Karim Foundation in Cambridge. Others also shared their experiences of navigating rural and predominantly White spaces, and their strategies for claiming space, giving voice to community members and creating opportunities for other culturally and racially diverse women leaders. One notable theme was the pathbreaking role each panellist had assumed during her career, opening (or at times creating) doors for herself and those who followed. Discussions across the day dwelled on temporality and the changing political landscapes as well as place-based challenges linked to rurality.

Overall, the leadership model, including this event and others and the growing leadership CoP, with a current membership of 25 emerging women leaders, aims to address these contextual and temporal challenges by bolstering and diversifying racialised women’s leadership in public services to ensure equity for minoritised victims-survivors.

By laying the groundwork for a regional advisory board through the CoP, the model challenges the epistemic erasure of racially and culturally minoritised women within the design of DASV, and broadly, public services which exacerbate the postcode lottery of services across England and Wales. Inspired by Pawson and Tilley’s (1997)[1] approach to realist evaluation (which seeks to understand what works for whom and in which circumstances) Mirna and Katherine aim to evaluate the impact of the leadership programme and CoP on how emerging leaders navigate specific challenges rooted in specific professional, relational and spatial contexts.

[1] Pawson, R. and Tilley, N., 1997. Realistic evaluation.

For further information, please contact Mirna at mirna.guha@aru.ac.uk

Photograph courtesy of Dr Mirna Guha and Dr Katherine Allen.

VISION member awarded UKDS Impact Fellow to study violence and mental health impacts in older age

VISION Research Fellow, Dr Anastasia Fadeeva, has been awarded a UK Data Service (UKDS) Fellowship.

Anastasia’s interest and education in medicine increased her awareness of the impact of social determinants on people’s health. This led to an MSc in Public Health at London Metropolitan University followed by a PhD at Northumbria University and a career in health services and public health research.

As a UKDS Fellow, Anastasia will look at the issues of violence in older age, the long-term impacts of violence on mental health, and the lack of reliable data.

For more information about Anastasia and her work, see her blog on the UKDS website or email her at anastasia.fadeeva@citystgeorges.ac.uk.

The UKDS is funded by the UKRI and houses the largest collection of economic, social and population data in the UK. Its Data Impact Fellowship scheme is for early career researchers in the academic or the voluntary, community, and social enterprise (VSCE) sector. The purpose of the programme is to support impact activities stemming from data-enhanced work. For further information on the UK Data Service please see: UK Data Service

Photograph from Adobe Photo Stock subscription

United to End Violence Against Women and Girls: An Online Animated Campaign

Violence against women and girls (VAWG) is a pressing issue in Iran, a Middle Eastern country marked by its patriarchal structure and systematic and pervasive gender discrimination. Educational programmes addressing this issue are scarce, and cultural barriers often hinder open discussion. The United to End Violence Against Women and Girls campaign aims to break this silence through a series of animated videos in Farsi and English and images designed to inform public discourse and to empower victims to seek support.

The United to End Violence Against Women and Girls project was led by VISION researchers Ladan Hashemi and Sally McManus, in collaboration with colleagues from other UK universities including the University of Bristol, Goldsmiths University, Animation Research Centre at the University for the Creative Arts, and Leeds Beckett University.

They worked with an animation production team in Iran, a social media advisor, and two advisory groups. The advisory groups were Mehre Shams Afarid, an Iran-based non-governmental organisation (NGO), and IKWRO, a London-based charity providing services to women victims of violence from the Middle Eastern and North African (MENA) region—to incorporate culturally specific insights.

Although the project initially focused on Iran, engaging with the UK-based NGO revealed an interest in extending its reach. As a result, English subtitles were added to make the animations accessible to a wider audience. This collaboration helped the content resonate with audiences both in Iran and within the global diaspora community, particularly those from the MENA region.

The animations are grounded in evidence from a survey of 453 women in Iran, which explored the manifestation of various forms of VAWG in Iran and women’s perspectives on how to eliminate it. The survey was designed by Fatima Babakhani, CEO of Mehre Shams Afarid.

Key findings from participants’ open-ended responses to the survey showed that, despite structural inequalities and deeply ingrained societal, cultural, and religious norms that perpetuate VAWG, change is possible through education and legal reforms.

As one survey participant noted: “Unfortunately, many still don’t understand what violence truly is. Raising awareness is the solution.”

The first four United to End Violence Against Women and Girls campaign animations focus on coercive control, economic abuse, technology-facilitated abuse, and active bystander interventions, with two more animations in development.

With guidance from an Iranian social media advisor, a digital strategy was developed to maximise the campaign’s impact. Instagram was chosen as the primary distribution platform, as it is the most widely used social media platform in Iran, with over 47 million users. The animations are also shared on YouTube to further extend the campaign’s reach.

Influencers and women’s rights activists with followings from thousands to millions were partnered with to amplify the campaign’s reach. The online campaign officially launched 25th November, on the International Day for the Elimination of Violence Against Women and Girls.

By leveraging evidence-based content and strategic partnerships, we hope to spark meaningful conversations and drive change across Iran and the diaspora communities from the MENA region.

Join us in raising awareness and advocating for change. Please follow and share the campaign links on your social media to help spread the message.

Link to Instagram page

Link to YouTube channel

This project was funded by City St George’s, University of London Higher Education Impact Fund (HEIF) Knowledge Exchange and by the UKPRP VISION research consortium.

For further information, please contact Ladan at ladan.hashemi@city.ac.uk

Discovering the Potential of Large Language Models in Social Science Research: Takeaways from an Oxford Workshop

By Dr Maddy Janickyj, Research Fellow in Natural Language Processing (NLP) for the Violence, Health, and Society (VISION) Consortium, University College London

As a data-focused VISION researcher with a PhD specialising in Natural Language Processing (NLP; see our previous blog for more about this), I initially avoided ChatGPT and similar tools. ChatGPT, a type of Large Language Model (LLM) developed by OpenAI, offers capabilities like summarising information, translating text, and even coding.

While ChatGPT is potentially the most well-known example of a LLM, similar models are integrated into many everyday tools. For instance, LLMs are the underlying technology in many customer service chatbots, virtual assistants like Alexa, and writing tools such as Grammarly. These LLMs are trained on large sets of data with the intention of getting them to understand (and in some cases generate) language. The models draw on this training to complete various tasks and are finetuned to work for specific domains. Their breadth of abilities and the many open-source models that have been developed make them the perfect methodological tool for researchers in both computer science and the social sciences. For clarity, an open-source LLM is one whose code and architecture are publicly available.

To further understand how LLMs are being used by researchers and to consider how the tools would integrate with and support violence-related research, I – a mathematician turned computational social scientist – attended the Oxford LLMs workshop. The event, held at Oxford’s Nuffield College, aimed to bring early-career scholars up to speed with the technical foundations, real-world applications, and research potential of LLMs. Throughout the week, I met with PhD/Masters students and other Post-doctoral researchers interested in using LLMs to evaluate anything from economic, linguistic, and political issues, for example.

Understanding LLMs: Lectures and Industry Insights

The first few days provided foundational lectures and talks, showcasing the technical underpinnings and application of LLMs. One of the big draws was the calibre of speakers. We heard from industry experts working at well-known companies such as Meta, Ori , Qdrant, Wayfair, Intento, Arize AI, and Google.

We then started our deep dive into LLMs, including how they are trained and evaluated. We heard about the numerous ways you can fine-tune LLMs, a step which occurs after general pre-training and tailors a model to meet domain/task-specific needs. Fine-tuning methods such as Continued Pre-training, Supervised Fine-tuning, and Preference Tuning were highlighted. Each technique offers different ways of adapting LLMs to specialised domains without needing to re-train them from scratch, saving computational resources.

We also covered common challenges associated with finetuning models. One of these is “catastrophic forgetting,” where a model’s performance declines in one area when it’s fine-tuned on another. For example, if a model is adjusted to improve name recognition, it may inadvertently lose accuracy in identifying locations. This side effect is something I encountered when finetuning other NLP models during my PhD and illustrates the balance required when refining LLMs.

Applying LLMs: Collaborative Research Projects

In the latter half of the week, workshop attendees collaborated on research projects, exploring LLM applications across social science realms. This was a hands-on opportunity to test LLM methodologies discussed earlier and apply them to real-world social science challenges.

Leading up to the workshop we had the chance to review the proposed project briefs, gather literature showcasing how LLMs are used in our respective disciplines, and finally rank the four projects according to our own skillsets and research interests. One of the projects we decided to tackle as a group focused on developing an LLM purely for social science research. LLMs are considered to have some form of bias, for example against certain demographic groups, and with this ongoing project, we wanted to create a fair, unbiased, and open-source LLM suited to the social sciences.

In another project, we examined gender bias in academia. For this, we used Google’s Gemini to classify the gender of authors in academic syllabi. By experimenting with prompts, we measured how well the LLM could assess gender trends in syllabus authorship. Using tools like Google Colab, we collaboratively coded and refined our approach, leveraging Gemini’s capabilities to highlight gender disparities effectively. In some cases, we found the model to correctly classify 100% of the authors’ genders. This project underscored both the potential and the limitations of LLMs in accurately capturing nuanced social issues.

Appreciating the Potential: Be Cautious

Overall, the Oxford workshop demonstrated how LLMs can be powerful tools in social science research including violence-related research such as what we do at VISION, provided they are tailored to specific domain needs and applied with caution. Hearing directly from researchers and industry professionals offered invaluable guidance on both leveraging and responsibly implementing LLMs. Its also important to consider the data you are utilizing and the outputs you are expecting. In my current area (which focuses on technology-facilitated abuse), an increasing number of researchers are using sensitive data and the outcomes of such research can impact the lives of real individuals. Thus, for anyone in the social sciences looking to integrate cutting-edge NLP methods, understanding the complexities behind these models and their applications is essential. I encourage readers to look at the work being done currently by the workshop participants, and to keep an eye out for later outputs of the workshop!

For further information, please contact Maddy at m.janickyj@ucl.ac.uk

Prior homelessness and associations with health and violent victimisation

By Dr Natasha Chilman, Research Associate, UKRI Population Health Improvement (PHI-UK), Population Mental Health Consortium, Kings College London

In the United Kingdom, we have the highest rate of homelessness compared to other high-income countries. For many people homelessness is a temporary, although often very impactful, experience in their lives. However, there is a paucity of research and data looking at people who are formerly homeless and living in private households (i.e., rented or owned accommodation).

This blog describes a new study which fills this gap, conducted by Dr Natasha Chilman from King’s College London and colleagues, including Professor Sally McManus from VISION.

The study authors analysed data from the Adult Psychiatric Morbidity Surveys, which is a nationally representative survey of adults living in private households in 2007 and 2014. Out of 13,859 people, 535 people reported previous experience of homelessness.

Some of the key findings of the study were:

A staggering 40% of people who formerly experienced homelessness had experienced violence in their homes at some point in their lives, compared to 7% of people who had never been homeless.
A quarter (24%) of people who formerly experienced homelessness reported experience of sexual abuse, compared to less than 5% of people who had never experienced homelessness.
Almost half (45%) of the formerly homeless group were currently experiencing depression or anxiety, compared to just 15% of people who had never experienced homelessness. People who formerly experienced homelessness were also experiencing more severe symptoms of these common mental disorders.
There were strong associations between former homelessness and health conditions, across common mental disorders, physical health conditions, alcohol/substance dependence, and multimorbidities. These associations persisted even after adjusting for a range of potential confounders, including indicators of socio-economic position and smoking.
Adjusting for adverse experiences including violence and abuse attenuated associations between former homelessness and alcohol/substance dependence related health outcomes, but not mental/physical health.

The findings from this study highlight the urgent need for long-term integrated healthcare support for people who are formerly homeless to continue after they have secured private housing. There were severe inequalities in experiences of violence and sexual abuse for people who have experienced homelessness, underscoring the importance of both violence and homelessness prevention, and of trauma-informed approaches to support.

To read the article or download the paper free of charge:

The public health significance of prior homelessness: findings on multimorbidity and mental health from a nationally representative survey | Epidemiology and Psychiatric Sciences | Cambridge Core

To cite:

Chilman N, Schofield P, McManus S, Ronaldson A, Stagg A, Das-Munshi J. The public health significance of prior homelessness: findings on multimorbidity and mental health from a nationally representative survey. Epidemiology and Psychiatric Sciences. 2024;33:e63. doi:10.1017/S2045796024000659

Or for further information, please contact Natasha at natasha.chilman@kcl.ac.uk

Empowering voices against ‘honour’-based abuse: A call to action

There is an urgent need for specialist support for Middle Eastern, North African (MENA) and Afghan women and girls living in the UK, with many facing the particular risk of ‘honour’-based abuse (HBA). On 18 October, the women’s rights organisation IKWRO held the impactful “Celebrating Courage: Empowering Voices Against Honour-Based Abuse” conference. It was hosted at City St George’s and proudly supported by the Violence and Society Centre (VASC) and VISION consortium.

The powerful event featured art, presentations, panels, film, and spoken word performances that shed light on the often-overlooked experiences of ‘honour’-based abuse (HBA) and the urgent need for education and prevention.

HBA disproportionately affects women and girls, it frequently goes unrecognised and is often conflated with domestic violence. Through this event, IKWRO, VASC and VISION aimed to illuminate the unique aspects of HBA and foster a deeper understanding of its implications.

A diverse group of speakers, including survivors, advocates, lawyers, and researchers, shared their insights and experiences. This included powerful testimonies from victims/survivors, bravely recounting their experiences of HBA. Their voices resonated throughout the room, creating a sense of solidarity and support, while also providing a safe space for discussion.

Over 100 people attended. A conference highlight was the strong presence of representatives from police and central government. Their attendance demonstrated a commitment to addressing HBA and offered them and other attendees a unique opportunity to engage directly with key figures at the forefront of the fight against HBA.

One salient moment during the conference occurred when VISION researcher Dr Ladan Hashemi, during the discussion with the police officers regarding enhancements to law enforcement responses, emphasised that “violence constitutes violence, and abuse constitutes abuse, regardless of an individual’s identity or background. It is essential for law enforcement agencies to acknowledge their obligation to prevent and mitigate HBA in the United Kingdom”.

The event also featured innovative approaches to communicating the issue of violence through art and spoken word. These forms of expression not only captivated the audience but also fostered a deeper understanding of the complexities surrounding HBA. By bringing art into a conference setting, IKWRO, VASC and VISION wanted to highlight the positive impact of creative communication, showcasing the experiences of survivors and the urgent need for change.

In honour of the International Day of the Girl Child, this event was an important step toward raising awareness about HBA and the specific challenges faced by women and girls in the UK to police, government, specialist services, academic researchers and the general public. A world free of abuse and violence requires the prevention of HBA, and for IKWRO that starts in their home base of London.

Key to the event was the organisation and support of VISION’s Knowledge Exchange Manager, Kimberly Cullen, and Dr Hashemi. IKWRO, VISION, and VACS will continue to work collaboratively to elevate the voices of survivors and advocate for meaningful change in the fight against ‘honour’-based abuse.

Photo caption: IKWRO and VISION. Photograph supplied by IKWRO.

Green space may be important in the prevention of crimes

The United Nations (UN) Sustainable Development Goals such as Goal 16, Promote peaceful and inclusive societies for sustainable development, provide access to justice for all and build effective, accountable and inclusive institutions at all levels, highlight the importance of using policy tools, for example urban planning, to prevent crimes. However, existing evidence of the association between green space and crime is mixed. Some studies indicate that the inconsistencies may be due to the variance in types of vegetation and the rates of crime reported across regions and countries.

Therefore, UK Prevention Research Partnership funded consortia, GroundsWell and VISION, worked together to assess the conditional association between green space and crime. Groundswell researchers Dr Ruoyu Wang, Dr Claire L. Cleland, Dr Agustina Martire, Prof Dominic Bryan, and Prof Ruth F. Hunter collaborated with VISION researchers Dr Ruth Weir and Prof Sally McManus to consider the influence of vegetation type such as grassland and woodland, crime type such as violence and theft, and the rates of crime reported in Northern Ireland.

They found that the association between green space and crime varies by vegetation type, crime type and rates of crime. The analyses showed that relatives were:

More grassland may be associated with lower crime rates, but only in areas with relatively low crime rates.
More woodland may also be associated with lower crime rates, but only for areas with relatively high crime rates.
The associations between green space and crime varied by type of crime.

Check out their recent publication, Rethinking the association between green space and crime using spatial quantile regression modelling: Do vegetation type, crime type, and crime rates matter?, where they discuss their findings further as well as the implications for government approaches to consider green space as a potential crime reduction intervention. Policymakers and planners should consider green space as a potential crime reduction intervention, factoring in the heterogeneous effects of vegetation type, crime type and crime rate.

To read the article or download free of charge:

Rethinking the association between green space and crime using spatial quantile regression modelling: Do vegetation type, crime type, and crime rates matter? – ScienceDirect

To cite:

Wang, R., Cleland, C. L., Weir, R., McManus, S., Martire, A., Grekousis, G., Bryan, D., & Hunter, F. R. (2024). Rethinking the association between green space and crime using spatial quantile regression modelling: Do vegetation type, crime type, and crime rates matter?. Urban Forestry & Urban Greening.

Illustration / photograph licensed under Adobe Photo Stock

The story so far: Co-production in Lambeth

By Elizabeth Cook, Senior Lecturer in Criminology & Sociology at City St George’s, University of London

As the VISION consortium approaches the end of its third year, work continues on consolidating the learning from various large datasets in crime and justice, health, and specialist services.

What we know is that these datasets are structured in different ways, collected by different agencies, and curated for quite different purposes. They represent particular ways of knowing about violence and abuse: they can help to identify patterns (e.g., what determines whether victim-survivors of sexual violence and abuse access support), prevalence (e.g., of workplace bullying and harassment), trends over time, and associations (e.g., between intimate partner violence, suicidality, and self-harm). However, we also know that large datasets struggle to capture the complex, and sometimes messy, realities of violence and abuse experienced by communities, especially those that are marginalised and minoritised.

Peer action research in Lambeth

In Lambeth, working in collaboration with peer researchers has made visible the evidence gaps that emerge at the intersection of multiple systems of inequality, including racism and misogyny.

We are lucky to be partnered with Lambeth Peer Action Collective (LPAC), High Trees and Partisan as part of a peer action research project. The aim of the project is to explore the role that trusted adults and trusted spaces can play in protecting young people from exposure to violence. Currently, there are 11 peer researchers that work as part of the LPAC: a collective of young people and youth organisations campaigning for change in their community. They are supported by High Trees, a Community Development Trust in Tulse Hill, eight partner youth organisations, and Partisan, a Black-led Community Interest Company providing culturally sensitive mental health support.

What has been achieved so far?

The project builds upon research conducted by the previous cohort of LPAC researchers conducted between December 2021 and August 2022. This project identified the impacts of violence on young people in Lambeth and the structural conditions of poverty, housing, education, urban regeneration, and public safety that were experienced unequally across the community.

Developing these findings further, the second cohort of peer researchers have been participating in weekly research training sessions led by High Trees and supported by VISION. The group has been learning everything they need for the next stage: from safeguarding and finances, to developing research questions, critical thinking skills, and how to evaluate research methods. This month, the LPAC researchers are getting ready to put into practice the interview skills that they have been learning each week in preparation for the next stage of the project – recruitment.

There has been amazing progress so far – not only in forming a research question and defining key concepts, but in developing a shared space for researchers to feel like change is possible and to collaborate with others who want the same.

What have we learned?

There are ongoing conversations about how peer action research can work to redress the imbalance between ‘researcher’ and ‘researched.’ These conversations seem even more relevant to research on violence and abuse, where the issue of power is central to both.

So far, the weekly sessions with peer researchers as well as our meetings with High Trees have taught us a lot about how power operates within institutions and the ways that it can be shared if there is a will to share it. This can be reflected in adequate resourcing, decision-making, access, and sharing skills and knowledge. The project has underlined the importance of respect in research: for different forms of expertise, within spaces, and within research relationships. The project has also challenged adult-centric assumptions about what we suppose that young people need to live better lives.

As mentioned previously, this project highlights the evidence gaps that occur at the intersection of multiple inequalities. In doing so, peer action research can also shape how we utilise large datasets, recognising how different social realities are reflected within existing data (or not).

In this sense, this collaboration has also made hyper-visible the question of: what and who is research for? As others have suggested, action research is not so much a methodology, but a way of thinking about research: it is a way of approaching a specific problem through community, participation, and curiosity. It is not necessarily driven by knowing more about something, but by wanting to change something with what you know.

We hope that this research continues in that spirit!

Further information

Do check out the LPAC’s manifesto for change and their previous report!

Photograph is copyrighted to Lambeth Peer Action Collective and not for use.

Natural Language Processing: Improving Data Integrity of Police Recorded Crime

By Darren Cook, Research Fellow in Natural Language Processing at City, University of London

Did you know that police recorded crime data for England and Wales are not accredited by the UK’s Office for Statistics Regulation (OSR)? This decision, made by the OSR after an audit in 2014, was due to concerns about the reliability of the underlying data.

Various factors affect the quality of police-recorded data. Differences in IT systems, personnel decision-making, and a lack of knowledge-sharing all contribute to reduced quality and consistency. Poor data integrity leads to a lack of standardisation across police forces and an increase in inaccurate or missing entries. I recently spoke about this issue at the Behavioural and Social Sciences in Security (BASS) conference at the University of St. Andrews, Scotland.

Correcting missing values is no small feat. In a dataset of 18,000 police recorded domestic violence incidents, we found over 4,500 (25%) missing entries for a single variable. Let’s assume it takes 30 seconds to find the correct value for this variable – that’s 38 hours of effort – almost a full working week. Given that there could be as many as twenty additional variables, it would take over four months to populate all the missing values in our dataset. Expanding such effort across multiple police forces and for multiple types of crime highlights the inefficiency of human-effort in this endeavour.

In my talk, I outlined an automated solution to this problem using Natural Language Processing (NLP) and supervised machine learning (ML). NLP describes the processes and techniques used by machines to understand human language, and supervised ML describes how machines learn to predict an outcome based on previously seen examples. In this case, we sought to predict the relationship between the victim and offender – an important piece of demographic information vital to ensuring victim safety.

The proposed system would use a text-based crime ‘note’ completed by a police officer to classify the victim offender relationship as either ‘Ex-Partner”, “Partner”, or “Family” – in keeping with the distinction made by Women’s Aid. Crime notes are an often-overlooked source of information in police data, yet we found they consistently referenced the victim-offender relationship. The goal of our system, therefore, was to extract the salient information from the free-form crime notes and populate the corresponding missing value in our structured data fields.

Existing solutions based on keywords and syntax parsing are used by multiple UK police forces. While effective, they require manual effort to create, update, and maintain the dictionaries, and they don’t generalise well. Our supervised ML system, however, can be automatically updated and monitored to maintain accuracy.

When tested, our system achieved 80% accuracy, correctly labelling the relationship type in four out of five cases. In comparison, humans performed this task with approximately 82% accuracy – an arguably negligible difference. Moreover, once trained, our system could classify the entire test set (over 1,000 crime notes) in just sixteen seconds.

However, we noted some limitations, the biggest of which was a high linguistic overlap in crime notes between ‘Ex-Partner’ and ‘Partner’ that caused several misclassifications. We believe more advanced language models (i.e., word embeddings) will improve discrimination between these relationships.

We also discovered a potential prediction bias against minorities. Although victim ethnicity wasn’t included in our training setup, we observed reduced accuracy for Black or Asian victims. The source and extent of this bias are subjects of ongoing research.

Our findings highlight the promise of automated solutions but serve as a cautionary tale against assuming these systems can be applied carte blanche without careful consideration of their limitations. Several outstanding questions remain. Is a system with 80% accuracy good enough? Is it better to leave missing values rather than predict incorrect ones? Incorrectly identifying a perpetrator as a current partner rather than an ex-partner could significantly impact the victim’s safety. Additionally, a model biased against certain ethnicities risks overlooking the specific needs of minority groups.

The conference sparked lively and engaging conversation about many of these issues, as well as the role that automation can be play within the social sciences more broadly. A research article describing these results in full is the focus of ongoing work, and the presentation slides are available below as a download.

For further information please contact Darren at darren.cook@city.ac.uk or via LinkedIn @darrencook1986

Dr Darren Cook, An application of Natural Language Processing (NLP) to free-form Police crime notes – 1 download

D Cook_NLP and Police Download

Photo by Markus Spiske on Unsplash