Your search

This practical guide from the UK Evaluation Society explores how evaluators can use AI tools responsibly and transparently across all stages of the evaluation process.

Earlier this year we spoke to 102 leaders from across the globe to ask: Ten years from now, what will we regret not having done today? This framing turned out to be a powerful device for collective horizon scanning and foresight. Across every conversation there was a sense that this moment demands more from us all, along with a strong eagerness to explore what we can own and act on. More than that, these discussions surfaced direction, hope and a way forward. From CEOs, professors, PhDs,...

This paper shares an insider's perspective on the spread of Thinking and Working Politically (TWP) across the US Agency for International Development (USAID), highlighting lessons learned along the way that are relevant even beyond the dissolution of the Agency in 2025.

The CGIAR 2030 Research and Innovation Strategy commits organizational change with seven ways of working, including “Making the digital revolution central to our way of working”. In that context, Artificial Intelligence (AI), introduces both opportunities and risks to evaluation practice. Guided by the CGIAR-wide Evaluation Framework, integrating AI tools requires a governance approach to balance innovation with ethical responsibility, ensuring transparency, fairness, accountability, and...

One of the main questions evaluators have when using AI is: "How can I know if it performs well?" To answer this question, we have developed a guidance note aimed at capturing what we have learned so far about integrating AI into evaluation and to offer a framework for further exploration. Since we began our experiments with Large Language Models (LLMs) in the Spring of 2023, we have made significant progress in using Generative Artificial Intelligence (GenAI) for processing and analyzing...

Generative AI (GenAI) is evolving rapidly, and its use in violence against women research remains new and largely experimental. While researchers are beginning to explore GenAI’s potential, uncertainty around its risks, ethical challenges, and real-world application persists. Recognising the need for clear guidance, the Sexual Violence Research Initiative (SVRI) partnered with The MERL Tech Initiative (MTI) to develop this resource. This guide aims to provide background on GenAI, strengthen...

This paper describes, synthesises and analyses the work of 25 Action Research Groups that were the heartbeat of the CLARISSA Systemic Action Research programme. The CLARISSA programme worked to create participatory and systemic solutions to the worst forms of child labour in the adult entertainment sector in Nepal, and the leather sector in Bangladesh. The paper explores how the action research groups were set up, and how they worked through key stages of trust and relationship building,...

This five-step framework, developed and tested by a foundation, embeds learning in emergent systems change strategies. It prioritizes the testing of hypotheses and assumptions, uses learning questions, and calls for examining both confirming and disconfirming evidence. --- A framework for embedding learning in systems change strategies and for testing strategic uncertainties. Learning and evaluation approaches that accompany systems change efforts need to fit with and support the emergent...

The Reduced Access Analytical Methods (RAAM) toolkit is a practical resource designed to help humanitarian practitioners overcome monitoring challenges in reduced access environments. Reduced access can be caused by natural disaster, conflict, political instability, or other factors, and typically makes it difficult to conduct normal monitoring of program implementation. The RAAM toolkit offers technical and managerial tools for a menu of analytical methods that can fill information gaps...

Over the course of this paper we lay out the basis of the Carve-Out method, an approach that leverages a behavioural framework to allow organisations (even those burdened with layers of bureaucracy and entrenched ideas) to create intentionally designed environments that give people the space, resources and support they need to explore and test new ideas. Ideas that may go on to transform the core of their organisation. The Carve-Out method is based on a simple insight: That if we want to...

The guide begins with a section on what we at TAF have learned about practicing Strategy Testing, highlighting things that we didn’t know when we started out. Some of these lessons may seem obvious, but others may strike you as less so. In any case, we believe that, when taken together, they provide a useful reminder of some of the pitfalls that arise in efforts to practice Adaptive Management. Moreover, they help to remind us that there are multiple layers to doing Strategy Testing...

Child Labour: Action-Research-Innovation in South and South-Eastern Asia (CLARISSA) is an evidence and innovation-generation programme funded by the United Kingdom’s Foreign, Commonwealth & Development Office (FCDO), responding to the challenge of the worst forms of child labour (WFCL) in Bangladesh and Nepal. It is a challenge characterised by a poor understanding of its drivers and a lack of evidence on what works to combat it. To handle such fundamental uncertainty, the programme adopts a...

When social accountability interventions scale up and their sustainability depends on the interactions of many agents and system components, related results are rarely observable at the end of an intervention. The 2019 OECD Development Assistance Committee’s (OECD DAC) revamped evaluations criteria for assessing sustainability acknowledges that such results are often emergent, and should be monitored and evaluated with this in mind. It therefore emphasizes a turn towards assessing complex...

In early 2021 we were asked to support two initiatives in South-East Queensland by undertaking a developmental evaluation of how they collaborated across multiple partners to reduce, prevent and ultimately end homelessness. The initiatives, Brisbane and Logan Zero, started with the ‘Advance to Zero’ methodology, and adapted this as they progressed. We were asked to track how contributors collaborated to tackle the systemic nature of this challenge. This gave us the opportunity to work with...

The Covid-19 Responses for Equity (CORE) programme was a three-year initiative funded by the Canadian International Development Research Centre (IDRC) that brought together 20 projects from across the global South to understand the socioeconomic impacts of the Covid-19 pandemic, improve existing responses, and generate better policy options for recovery. The research covered 42 countries across Africa, Asia, Latin America, and the Middle East to understand the ways in which the pandemic...

CLARISSA (Child Labour: Action-Research-Innovation in South and South-Eastern Asia) has a participatory and child-centred approach that supports children to gather evidence, analyse it themselves and generate solutions to the problems they identify. The life story collection and collective analysis processes supported children engaged in the worst forms of child labour in Bangladesh to share and analyse their life stories. Over 400 life stories were collected from children who worked in...

Organising for Systems Innovation at Scale Our team at Griffith Centre for Systems Innovation have been experimenting with and evolving a Challenge-led Innovation Approach (based on Mission-oriented approaches developed by Mariana Mazzucato at UCL IIPP and others internationally). We are using this approach to guide the way we work internally and engage with our systems innovation partners. We’ve facilitated intensive Re:Treats, worked with government bodies, businesses and civic...

The RoL method is a visual narrative method that helps people tell stories of the past, present, and future. Individuals can use this method to introduce themselves in a fun and descriptive way. A group can use it to understand and reflect on the past and imagine the future of a project. Besides, it can also be used to build a shared view of a process over time while acknowledging different and perhaps contradictory perspectives. The method uses drawings rather than text, making it useful in...

This report addresses the well-recognized evidence gap1 on the longer-term impacts created by marketdriven programming; specifically, programming influenced by market systems development (MSD) principles. It does so by presenting the findings of an ex-post study conducted three and a half years after the close of USAID’s Feed the Future Senegal Naatal Mbay Activity (hereafter Naatal Mbay) in 2019. It examines the scale and sustainability of changes resulting from Naatal Mbay’s introduction...