Typologies You Can Count: A Rule-Based Caregiver Typology Linking Photovoice Evidence to Panel Monitoring Data

Small-scale community organizations frequently generate bifurcated evidence streams: localized,

epistemologically rich qualitative datasets and structured, quantitative monitoring panels. This

paper introduces an operational, low-cost methodology to mathematically bridge this divide.

We adapt recent persona-quantification frameworks from digital-inclusion research to evaluate

unpaid care work (UCW) within forest-adjacent communities in Kole District, Northern Uganda.

Utilizing inductive photovoice research (n= 15) and a deliberative community dialogue (n= 173),

we derive five discrete caregiver configurations based on workload density, care-redistribution

dynamics, and enterprise engagement. We formalize these configurations as ordered, deterministic,

piece-wise logical rules executed over a longitudinal household monitoring panel (n= 80) via

KoboToolbox. This framework allows individual households to be typified dynamically each

quarter, maps household trajectories through an empirical transition matrix, and calibrates

localized sample prevalence against national census data to size district-level policy interventions.

We demonstrate that rule-based taxonomies offer a more statistically honest and politically

legible alternative to unsupervised algorithmic clustering for small-scale community interventions,

and we explicitly delineate the boundary conditions of this approach.