Contact Us
1. Overview of the Most Significant Change (MSC) Technique
The Most Significant Change (MSC) technique is a form of participatory monitoring and evaluation that relies on the collection, discussion, and systematic selection of stories describing change experienced by programme or project stakeholders, rather than relying primarily on predefined indicators of change (Davies and Dart, 2005; INTRAC, 2017). It is participatory in two distinct senses: stakeholders help decide what kinds of change should be tracked, and they are directly involved in analysing the resulting stories (Davies and Dart, 2005). It is a form of monitoring because it can run continuously throughout a programme cycle, and it contributes to evaluation because it produces data on outcomes and impact that can help assess how a programme is performing overall. Since it does not depend on indicators fixed in advance, MSC is especially suited to programmes whose outcomes are difficult to predict, vary widely across beneficiaries, or are the subject of disagreement among stakeholders about which results matter most (INTRAC, 2017; Davies and Dart, 2005).
MSC was developed during Rick Davies’ fieldwork in Bangladesh in the mid-1990s while working with the Christian Commission for Development in Bangladesh (CCDB), a Bangladeshi NGO with over 550 staff and a programme reaching more than 46,000 people across 785 villages (Davies, 1998; Davies and Dart, 2005). Davies developed the approach as part of his doctoral fieldwork on organisational learning, developing an alternative to conventional indicator-based monitoring in favour of asking participants what they judged to be the most significant change in their lives over a given period (Davies, 1998). He described the resulting design as an evolutionary approach to organisational learning, borrowing the logic of variation, selection, and retention from evolutionary theory: stories of change are generated at the field level, filtered through successive tiers of an organisation, and the winning accounts, together with the reasoning behind their selection, are documented and fed back down the hierarchy, refining the search for significant change in the next reporting cycle (Davies, 1998; Davies and Dart, 2005).
2. The Ten-Step Process
Davies and Dart (2005) set out ten steps for a full implementation of MSC, though they identify only three as strictly indispensable: collecting significant change stories, selecting the most significant among them through a designated panel, and feeding back the results of that selection to stakeholders. INTRAC (2017) condenses the process into five practical stages widely used as a quick reference: defining domains of change; deciding how and when to collect stories; collecting the stories; selecting the most significant stories; and verifying them. The Equal Access Nepal manual, developed by Lennie et al. (2011) for an Assessing Communication for Social Change project, adapts the same ten-step structure for use by field-based monitoring and evaluation staff training community researchers. The ten steps are as follows.
1. Raising interest and getting started. The first step generally involves introducing a range of stakeholders — project or programme staff, other levels of the organisation, targeted beneficiaries, and often donor representatives — to the technique, in order to build interest and a sense of ownership over the process (INTRAC, 2017; Davies and Dart, 2005). Davies and Dart (2005) recommend piloting MSC on a small scale before a full roll-out, and identifying “champions” who can excite and motivate people, answer questions, and help organise the collection and review of stories. Lennie et al. (2011) echo this advice for Equal Access Nepal, recommending a small, simple pilot with the most enthusiastic staff before wider implementation.
2. Defining domains of change. Stakeholders next identify “domains of change”: broad, deliberately loose categories, typically between three and five, within which significant events will be reported, rather than precisely defined performance indicators (INTRAC, 2017; Davies and Dart, 2005). In the original CCDB application, domains covered changes in people’s lives, changes in people’s participation, changes in the sustainability of people’s organisations and activities, and an open “any other change” category left deliberately unconstrained so that field staff could report whatever they judged important (Davies, 1998; Davies and Dart, 2005). Domains are routinely adapted to context: Lennie et al. (2011) shaped Equal Access Nepal’s domains around its community radio objectives, while Ghadirian et al. (2022) organised a Ghanaian nutrition-education evaluation around participant, peer, family, and school-level domains, plus an “other” category for shifts in health outcomes.
3. Defining the reporting period. Stakeholders decide how frequently stories will be collected; Davies and Dart (2005) note that this has ranged from fortnightly to yearly across different applications of MSC, with three-monthly reporting the most common. In the original CCDB case, an early trial of fortnightly reporting proved too demanding of staff time at both project and head-office level, and the programme settled on monthly reporting instead (Davies, 1998)
4. Collecting stories of change. Stories are collected from those most directly involved in a programme — typically participants and field staff — by asking respondents a version of the central MSC question. In the original CCDB case, this was worded: “During the last month, in your opinion, what do you think was the most significant change that took place in the lives of people participating in the PPRDP project?” (Davies, 1998). Respondents are also asked to explain why they consider that change the most significant (Davies, 1998; Davies and Dart, 2005).
5. Selecting the most significant stories. Groups of designated stakeholders read or listen to the stories aloud, discuss their relative value, and agree on the single most significant account within each domain. This is repeated at successive organisational levels — from community groups, to programme level, to organisation level, and sometimes on to funders — until a handful of stories represent the most significant changes overall (Davies and Dart, 2005; INTRAC, 2017).
6. Feeding back the results. The criteria used for each selection, and the reasoning behind it, are recorded and fed back down the hierarchy to storytellers and other stakeholders, so that subsequent rounds of story collection are informed by what previous reviewers valued. This feedback loop is considered a defining, indispensable element of MSC (Davies, 1998; Davies and Dart, 2005; Lennie et al., 2011; Dart, 2000).
7. Verification. Although optional, verification is treated as important for safeguarding the credibility of MSC: stories can be checked against the accounts of others involved, against site visits, or against other supporting evidence, so that inaccurate or misleading accounts are rejected before being passed further up the hierarchy (Davies and Dart, 2005; INTRAC, 2017).
8. Quantification. Also optional, quantification is achievable in at least three ways: recording basic statistics within an individual story, such as how many people benefited from a change; tallying how often a given type of change is mentioned across all collected stories, including those not selected; or asking other stakeholders whether they have observed or experienced a change similar to the one selected as most significant, which gives some indication of how widespread that change is (Davies and Dart, 2005; INTRAC, 2017).
9. Secondary analysis and meta-monitoring. The MSC process itself can be monitored — tracking who participates, how widely stories are drawn from across the organisation, and which kinds of change are reported most often (Davies and Dart, 2005). Davies (1998) found that CCDB’s system had generated accounts of change in 43 per cent of local associations (shomitis) within six months of implementation, a proportion that continued to grow.
10. Revising the system. The final step is to revise the design of MSC to reflect what has been learned from using it and from analysing its use, adjusting domains, reporting periods, or selection processes as the organisation’s needs evolve (Davies and Dart, 2005; Lennie et al., 2011).
3. When to Use the Most Significant Change (MSC) Technique
MSC is better suited to some programme contexts than others (Davies and Dart, 2005). According to Davies and Dart (2005) and INTRAC (2017), MSC delivers the most value in programmes that are complex and produce diverse, emergent outcomes; large, with numerous organisational layers; focused on social change; participatory in ethos; built around repeated contact between field staff and participants; or struggling to make sense of impact through conventional monitoring systems. It is most useful, INTRAC (2017) notes, where it is not possible to predict in detail what a programme’s outcomes will be, where outcomes vary widely across beneficiaries, where stakeholders disagree about which outcomes matter most, or where interventions are expected to be highly participatory. MSC is generally suited to monitoring focused on learning rather than accountability alone, and to situations where evaluators are keen to capture the effect of an intervention in participants’ own words (Davies and Dart, 2005).
Conversely, Davies and Dart (2005) caution that MSC may not justify its cost in simple programmes with easily defined outcomes, where quantitative monitoring is sufficient and considerably less time-consuming. There are often quicker ways to capture expected change, produce material for public relations, conduct a retrospective evaluation of a completed programme, understand the average experience of participants, or complete an evaluation report for accountability purposes alone, and MSC is not the most efficient tool for these specific tasks (Davies and Dart, 2005; INTRAC, 2017).
Successful implementation also depends heavily on organisational context. Davies and Dart (2005) and Lennie et al. (2011) identify closely overlapping enabling conditions: a culture in which it is acceptable to discuss things that go wrong as well as successes; champions with strong facilitation skills who can promote the technique; a willingness to try something different; sufficient time to run several cycles of story collection and selection; infrastructure to enable regular feedback of results to stakeholders; and commitment from senior management. Where these conditions are absent, both sources suggest MSC is unlikely to be sustained or to deliver reliable insight.
4. MSC within the Wider Field of Participatory Monitoring and Evaluation
MSC is one of several qualitative, story-based tools within the broader family of participatory monitoring and evaluation approaches, and Guijt (2014), writing for UNICEF’s Methodological Briefs on impact evaluation, presents it alongside other such techniques. Guijt (2014) cautions against equating “participatory” with “qualitative”: participatory methods, in her account, can just as readily be used to collect quantitative data, and MSC’s story-based format should not be taken as the only way participation can be built into monitoring and evaluation. She also cautions that participatory evaluation more broadly risks obscuring the power dynamics that determine whose information, and whose conclusions, are prioritised over others’.
Davies and Dart (2005) situate MSC more specifically relative to other forms of participatory monitoring and evaluation (PM&E), noting that MSC differs from many other PM&E approaches in that its data take the form of text-based accounts of change, and in the way it combines participatory analysis with a systematic, tiered selection process. They also observe that MSC differs from more ad hoc and egalitarian participatory processes in that it deliberately works through existing organisational hierarchies and power structures using established layers of authority to filter and select stories rather than trying to reach conclusions outside them.
Davies and Dart (2005) draw a related comparison with case-study and vignette methods, arguing that MSC’s distinguishing strength is that it makes explicit who selected each story, from how many alternatives, over what period, and why — information that conventional case studies rarely disclose. They also compare MSC with Appreciative Inquiry, noting that both approaches focus on what is working in order to encourage more of it, but that Appreciative Inquiry is oriented toward future visioning and is not necessarily an ongoing process in the way MSC is designed to be, and MSC uses structured, tiered selection processes that Appreciative Inquiry does not.
5. Applications
The flexibility of MSC is evident across the range of settings in which it has been applied. Its origin case remains CCDB’s rural development programme in Bangladesh, where an experimental nine-step monitoring system was piloted across four project areas in 1994 and had, within six months, generated accounts of change in 43 per cent of local associations, a proportion that continued to grow despite ongoing uncertainty about the system’s longer-term institutional future (Davies, 1998). In Australia, Dart (2000) adapted the method for the Target 10 dairy-extension project, running it as a continuous monitoring process from 1998 across four regions of Victoria; a subsequent meta-evaluation, based on a facilitated discussion with project funders, interviews with committee members and staff, and a staff survey, found that participants reported a stronger shared understanding of programme impact and, in some cases, a morale boost from seeing their contributions recognised in selected stories.
More recently, Lennie et al. (2011) documented the adaptation of MSC for Equal Access Nepal’s evaluation of two community radio programmes, incorporating it into a broader participatory monitoring and evaluation toolkit for field-based community researchers. Ghadirian et al. (2022) combined MSC with participatory video in a cluster-randomised trial of a nutrition-education intervention among adolescent girls in Ghana: after screening two rounds of participant-made videos, project staff collected 116 “headlines” of significant change, narrowed these to 14 through a nutrition-literacy selection rubric, and had a community advisory board select the four most significant stories, one per domain, revealing among other things how adolescents found creative ways to secure iron-rich foods and to influence household and school food practices. Taken together, these applications illustrate both the durability of Davies and Dart’s original design and the extent to which practitioners across sectors, including agricultural extension, community media, and public health nutrition, have adapted its domains, selection panels, and quantification steps to their own institutional needs.
6. Strengths and Limitations
MSC’s principal strength lies in its capacity to surface unexpected and unpredictable change in complex, participatory programmes where pre-defined indicators would either miss or misrepresent what matters most to stakeholders (INTRAC, 2017; Davies and Dart, 2005). It requires no specialised technical training, encourages analytical reflection because participants must justify why one change matters more than another, and generates thickly described, richly contextualised accounts rather than a single simplifying number (INTRAC, 2017; Davies and Dart, 2005; Lennie et al., 2011). Because it is transparent about who selected which stories and why, it also lends itself to organisational learning in a way many conventional monitoring systems do not (Davies and Dart, 2005; Dart, 2000).
Because MSC does not rely on statistical measures of significance, its proponents have had to make a separate case for its validity. Davies and Dart (2005) identify six mechanisms underpinning its credibility: thick description of events in their local context; a systematic, disciplined process of selection by designated panels; transparency about who selected which stories and why; verification against independent evidence; participation of a wide range of stakeholders; and member checking, whereby storytellers can review and correct the documented version of their account. Its sampling logic is explicitly purposive rather than representative: rather than characterising the average participant experience, it deliberately seeks out the most information-rich cases (Davies and Dart, 2005). This is consistent with Guijt’s (2014) broader observation that participatory evaluation processes, MSC included, need to be alert to the power dynamics that determine whose information and whose conclusions are prioritised over others’.
MSC is candid about its own biases. It tends to over-represent success relative to failure: in the Target 10 dairy-extension trial, about 90 per cent of stories concerned positive outcomes, and in one community-media programme in Laos the proportion ranged from 80 to 90 per cent (Davies and Dart, 2005). Both Davies and Dart (2005) and Lennie et al. (2011) note that this can be offset by designating a specific domain to capture negative change. Selection is also inescapably subjective, reflecting the values of whichever panel is doing the choosing, and can favour participants who are simply more skilled storytellers; both Davies and Dart (2005) and Lennie et al. (2011) treat this as a reason MSC should not be used as a stand-alone monitoring and evaluation tool, but should instead be combined with other methods. The process also privileges the voices of those who attend review sessions, which can under-represent stakeholders who do not participate — an issue of voice and power that can be mitigated through representative selection panels or parallel panels for different interest groups (Davies and Dart, 2005).
Against these strengths, MSC is not designed to capture typical or average experience, is resource and time-intensive if implemented properly, and is not well suited to reporting on finance, inputs, or activities (INTRAC, 2017; Davies and Dart, 2005). Its qualitative, purposively sampled data can sit awkwardly alongside donor demands for numerical reporting, though this can be partially addressed through the quantification strategies described in Section 2 (INTRAC, 2017). Ultimately, as Davies and Dart (2005) and INTRAC (2017) both stress, MSC is best understood not as a stand-alone evaluation methodology but as one tool within a wider participatory monitoring and evaluation repertoire (Guijt, 2014), whose value depends heavily on the rigour, transparency, and genuine stakeholder engagement that organisations are willing to invest in its implementation.
7. References
Dart, J. J. (2000). Stories for Change: A systematic approach to participatory monitoring. In Proceedings of Action Research & Process Management (ALARPM) and Participatory Action-Research (PAR) World Congress, Ballarat, Australia.
Davies, R. (1998). Professional practice: An evolutionary approach to facilitating organisational learning: an experiment by the Christian Commission for Development in Bangladesh. Impact Assessment and Project Appraisal, 16, 243-250.
Davies, R., & Dart, J. (2005). The ‘most significant change’(MSC) technique. A guide to its use, 10.
Ghadirian, M. Z., Marquis, G. S., Dodoo, N. D., & Andersson, N. (2022). Ghanaian female adolescents perceived changes in nutritional behaviors and social environment after creating participatory videos: a most significant change evaluation. Current Developments in Nutrition, 6(8), nzac103.
Guijt, I. (2014). Participatory approaches. Methodological Briefs: Impact Evaluation, 5(5), 2.
INTRAC. (2017). Most significant change. https://www.intrac.org/app/uploads/2017/01/Most-significant-change.pdf
Lennie, J., Tacchi, J., Koirala, B., Wilmore, M., & Skuse, A. (2011). The most significant change technique: A manual for M&E staff and others at Equal Access. Equal Access Participatory Monitoring and Evaluation Toolkit. BetterEvaluation. https://www.betterevaluation.org/sites/default/files/EA_PM%2526E_toolkit_MSC_manual_for_publication.pdf
AUTHORS
