Case stories

Projects in practice: the question, the approach, what was learned, and the limits of the evidence.

The design behind the work is described in Reinhard and Reddy (2026), Designing Sensemaking Studies, Palgrave Macmillan. How the geometry is read is on How the method works.

Our Tomorrows, Kansas

University of Kansas Center for Public Partnerships and Research.

Context

In 2019 Kansas received a Preschool Development Grant, and the project was built within it. The University of Kansas Center for Public Partnerships and Research (CPPR) built the Our Tomorrows project in three phases: statewide story collection, community sensemaking workshops, and community action labs (Flinders and Counts, 2024).

What the earlier method missed

The CPPR authors describe the limits of standard evaluation in their own terms: "Survey fatigue is real. Vulnerable populations are weary of extractive techniques without feedback loops." They also observe that the imagination needed to change a system is typically low among participants, and that abductive reasoning is uncommon in applied settings (Flinders and Counts, 2024). A survey tests the questions its designers already had. The project's purpose was to find what those questions had not named.

What was done

A single open prompt ("Remember a time when you felt like your family or another family you know was thriving, or just surviving…") was followed by triads that let participants characterise their own story without being forced into a single cause. Two examples of those triads: "views on raising a family were shaped by: personal experience / friends and family / tradition", and "who should have been responsible for making sure that kids thrived: the family / friends and community / government".

The team co-developed story-collection strategies with each partner organisation. The authors state that incorporating Liberating Structures and futures thinking into the sensemaking sessions is essential to move participants to a more generative state of mind (Flinders and Counts, 2024).

Reach, as reported in that paper: 2,279 stories from all 105 Kansas counties, submitted online, on paper, and through interviews by trained community collectors described there as citizen journalists.

Self-signification was aggregated into four themes (Bright Spots, Disruptors, Unmet Needs, Unheard Voices) and returned as regional story packs to fourteen two-hour Community Sensemaking Workshops. Participants sorted the stories using the participants' own interpretations. Community Action Labs then funded local experiments of up to $2,000. The authors report 46 action-lab proposals, double the number expected, funded through the Kansas Children's Cabinet and Trust Fund.

The State of Minnesota (MN StoryCollective) has since adapted the approach. Its policy-focused action labs bring agency staff and leaders together to work through community stories and consider changes.

An analytic reading, not the authors' own language, uses the CDE model. The statewide network of collectors and partner onboarding is the Container. The four themes and the contrasts between regions are the Differences. The story packs, workshops, and action labs are the Exchanges that carry patterns back to the people who can act. The paper itself does not use CDE language. This reading is imposed afterwards.

What it showed

The authors report examples of issues that surfaced from the stories and might not have appeared in a pre-set instrument: the cost of living through a pandemic without technology or reliable internet, the barriers facing licensed in-home childcare providers who rent their homes, and programme requirements that exclude vulnerable families.

A realist reading: context (a state-level mandate plus local onboarding) + mechanism (stories interpreted by their tellers, then by communities) = outcome (locally owned experiments and a story bank that the project's partners can act on).

Limits

The authors acknowledge that impact from this kind of work is hard to measure, that abrupt pivots and infrastructure challenges disrupted delivery, and that partners were often unsure how to begin. The story bank documents reach (105 counties) and not statistical representativeness. The paper does not describe a probability sample.

The paper describes the project as using more than one data platform, Spryng (spelled "Sprying.io" in the paper) among them, together with paper versions of the framework for in-person groups, and it does not report results platform by platform. The outcomes above are therefore attributed, as the authors attribute them, to the Our Tomorrows sensemaking approach, of which Spryng is one of the named data platforms (Flinders and Counts, 2024, pp. 364–376). The paper's count (2,279) differs from an OECD innovation-observatory description, which cites a higher figure. The peer-reviewed number is the one used here. The paper's "nearly 10,000 voices" refers to the wider programme, not to Kansas.

Queen's University and the International Organization for Migration

Gendered migration experiences of Venezuelan women and girls. Queen's University and IOM, funded by Elrha's Humanitarian Innovation Fund.

Context

A research collaboration between Queen's University and IOM, supported by Elrha's Humanitarian Innovation Fund, used a sensemaking approach to understand gendered migration experiences and gender-based-violence risks among Venezuelan refugee and migrant women and girls in Brazil, Ecuador, and Peru (Bartels and Noriega, 2022; Bartels et al., 2023).

What the earlier method missed

The topic is exactly where direct survey questions are weakest. The investigators listed the advantages they expected: indirect prompts; a richer range of responses than discrete options allow; reduced social-desirability bias, because no response option is more acceptable than another; reduced interpretation bias, because participants interpret their own experiences; and digital efficiency. These are the investigators' stated rationale, not an independent measurement of bias (Bartels and Noriega, 2022).

What was done

Enumerators invited participants, in public spaces such as border crossings, markets, transportation hubs, and points of service delivery, to record or type a story in response to an open prompt, then to plot their own interpretation on triads, with multiple-choice questions for demographic context. About 80% of narratives were audio-recorded.

Scale: the team aimed for 3,000–3,500 narratives over 10 weeks and obtained 9,339 narratives from 9,116 unique respondents over 12 weeks, across nine locations in three countries.

In July 2022 the team held seven workshops with 120 service providers and 14 focus groups with 116 Venezuelan refugees and migrants, including two groups of Indigenous participants and two of LGBTIQ+ participants, to share findings, test whether the framing was acceptable and aligned with lived experience, and brainstorm responses (Bartels et al., 2023).

An analytic reading, not the project's own language, uses the CDE model. The Container is the enumerator-led interview with trauma-aware training. The Differences are the triad axes and the contrasts between locations and groups. The Exchanges are the validation workshops and focus groups that fed findings back. The public reports describe the workshops and break-out groups without naming a particular facilitation method, and none is claimed here.

What it showed

The project report and the funder's account document reach and reception more fully than substantive findings. They show that a story-and-signifier method could reach thousands of people in transit on a sensitive topic and exceed its target roughly threefold. Respondents spanned ages (6.8% under 18; 43.3% aged 19–30; 34.8% aged 31–45; 13.8% over 45). The report's breakdowns include Indigenous, Afro-descendant, and LGBTIQ+ respondents and persons with a disability. Topics prioritised for analysis included child marriage and unions, transactional sex, human trafficking, the burden of care falling to women, mental health, and the experiences of LGBTIQ+ and Indigenous migrants. The report states that analyses were under way at the time of writing (Bartels et al., 2023).

Limits

The team reports noisy audio from outdoor settings; an emotional toll on enumerators (many were themselves refugees or migrants) that led to a second psychological-first-aid session and advice to reduce daily interview loads; a project start delayed by a year and a scope cut from five countries to three; and, because transcription and translation of all 9,339 narratives exceeded the budget, reliance on AI transcription and keyword search of English translations, with human verification only for narratives quoted in results. The authors note that keyword searching of translations inevitably misses some micronarratives and nuance (Bartels et al., 2023).

A separate Queen's University study of emergency care, published as Chard et al. (2025) in BMC Health Services Research, is a different project. It is not this case. https://doi.org/10.1186/s12913-025-12684-y

BC Employment and Assistance Appeal Tribunal

Context

The Employment and Assistance Appeal Tribunal (EAAT) is an independent tribunal that hears appeals from people denied income assistance, disability assistance, or child-care subsidies, many of them among the most vulnerable people in the province. Its Chair, Emily Drown, led an active-sensemaking project with outside consultants and partners. Her public post names Povnet, QED Insight, the Ministry of Social Development and Poverty Reduction, and law students from TRU, UBC, and UVic. The project received the Premier's Award in the Evidence-Based Design category in 2022 (Drown, 2022; EAAT annual reports; interview with Emily Drown).

What the earlier method missed

The tribunal mailed a user-experience survey with each final decision and got back "about one or two a year" — "so really they were telling us nothing" (interview: Drown, 2023). Leaders knew who phoned to complain but not what the people who did not call had experienced. A survey also asks a preconceived question. The Chair wanted to learn "what we didn't know we needed to ask about" (interview: Drown, 2023).

What was done

The Chair worked with the consultants to build prompts open enough to avoid leading people toward negative or positive stories. One example: "Tell us about a time you felt respected or disrespected at the tribunal." To test it, staff called a few recent appellants, who agreed to be "guinea pigs", and appeal coordinators who had worked with them asked them to complete it. The link was sent with a short paragraph from the Chair explaining why the tribunal wanted to know.

Data arrived almost at once. Within about five minutes of the invitation going out, appellants were writing that no one had ever wanted to know what they thought. Where people could not complete the instrument online, volunteer law students took stories by phone, acting as transcribers.

Because of the pandemic the whole staff worked in online workshops, grouping stories and metrics on shared whiteboards and creating their own titles for the stories. The Chair, initially sceptical ("a waste of staff, time and resources"), says that staff began suggesting process changes by the end of the first workshop (interview). Eighteen months later she reported bi-weekly staff round-tables on what each person had learned from appellants that week. The 2021/22 annual report records that the full complement of staff took part in multi-day workshops (EAAT, 2022).

An analytic reading, not the Tribunal's own language, uses the CDE model. The Container is the tribunal experience bounded in the prompt and the trust created by the Chair's invitation. The Differences are the contrasts the staff themselves surfaced when grouping stories. The Exchanges are the shared workshops that carried appellants' voices into staff decisions. The interview describes grouping, titling, and discussion in workshops without naming a particular facilitation method, and none is claimed here.

What it showed

Small details that a survey would not have asked about. The tribunal's practice of asking appellants, in front of the panel that will decide the appeal, to consent to a ministry observer was described in one story as a power dynamic the tribunal had not recognised. Another story showed that a hearing room overlooking a police station could aggravate a participant's trauma, and staff now consider what an appellant will see from the hearing-room window, with a trauma-informed lens.

The 2021/22 annual report lists changes that stemmed in part from the project: all written communications rewritten in plain language (most now read at a grade-4 level or lower), reformatted forms with diverse gender options, and a new accessible website. The 2022/23 report adds plain-language training for members, introductory calls to appellants, reminder calls before hearings, and reduced correspondence.

A realist reading: context (a vulnerable population, an agency that rarely heard from non-complainants) + mechanism (open story, self-interpretation, staff-wide sense-making) = outcome (a staff culture of continuous improvement and a package of small, low-cost changes).

Limits

The Chair herself describes the changes as "very, very small" and mostly cost-free, and the 2022/23 annual report states only that "anecdotal evidence suggests" they are making a large difference. No outcome measure is reported. The tribunal plans to repeat the project every three years to test empirically whether the changes remove barriers. The respondents were self-selected, and some appellants needed alternatives to online completion (for example, no computer access) or language support. In hindsight the Chair observed that phone-based capture may have suited a dispersed population better than gathering people in a workshop (EAAT annual reports; interview).

Sources

A practitioner’s perspective: keeping stories behind the patterns

It just gives me structure, that ability to point to a point on a graph or one of the triangles and say look, this is why it's there and show people the story that is behind it. That's very powerful.

— Dr. Robin Pharoah, anthropologist and advisor to Spryng, book interview, 11:00.

Robin is describing how he uses Spryng alongside ethnographic fieldwork: a structured view of the research that still lets him return to the individual account. This is his practitioner perspective, rather than an outcome measure for any one study.