Applied Implementation Research for Clean Cooking in Cambodia

Purchase will be defined as the acquisition of an induction stove during the trial period, as reported in household surveys and verified through study-linked sales records matched to household study IDs. The outcome will be calculated as the number of participating households with a confirmed induction stove acquired divided by the total number of households in each village. Results will be reported overall and stratified by household poverty status, verified by the Cambodian Government (IDPoor status).

Disuse will be defined as the reduction in biomass fuel use for cooking, measured as the proportion of meals cooked without biomass fuels (e.g., firewood and charcoal) during a defined recent reference period, as reported in household surveys. Results will be reported overall and stratified by household poverty status.

Daily electricity consumption (kWh) associated with induction cookstove use, measured using built-in, cloud-connected data loggers that record energy use during cooking events among households that purchase an induction cookstove and consent to monitoring. Results will be summarized at the household and village level and stratified by household poverty status.

Number of daily cooking events measured using built-in, cloud-connected data loggers installed on induction cookstoves. Cooking events will be derived from time-stamped stove activity among households that purchase an induction cookstove and consent to monitoring. Results will be summarized at the household and village level and stratified by household poverty status.

Total duration of induction cookstove use per day, measured using built-in, cloud-connected data loggers based on time-stamped stove activity among households that purchase an induction cookstove and consent to monitoring. Results will be summarized at the household and village level and stratified by household poverty status.

Stove stacking will be defined as the concurrent use of multiple cooking technologies within a household, including biomass (e.g., firewood and charcoal), liquefied petroleum gas (LPG), and induction stoves. Stove use will be measured using thermocouple-based temperature loggers (Geocene Dots) installed on non-induction stoves. These devices record stove temperatures continuously and transmit data via Bluetooth to mobile devices, which sync to a central server. The outcome will be calculated as the proportion of monitored days with no biomass stove use and summarized at the household and village level, with results stratified by household poverty status.

Personal exposure to PM2.5 and black carbon will be measured among primary cooks using wearable personal exposure monitors (Ultrasonic Personal Aerosol Sampler) placed in the breathing zone of the primary cook for a continuous 24-hour period. Measurements will be collected among households that purchase an induction cookstove and a time-series matched subset of households that do not purchase. The outcome will be defined as 24-hour average PM2.5 (µg/m³) and black carbon concentrations and summarized at the household and village level, with results stratified by household poverty status.

Induction cookstove use trajectories will be assessed as longitudinal patterns of use among households that purchase an induction cookstove (e.g., sustained, increasing, declining, or discontinued use), measured using built-in, cloud-connected stove data loggers, including frequency of use, duration of use, and electricity consumption. Results will be summarized across households and villages and stratified by household poverty status.

Determinants of sustained use, declining use, or discontinuation of induction stove use, as well as adoption and continued use of clean cooking technologies (e.g., LPG) and reliance on biomass fuels (e.g., firewood and charcoal), will be assessed using after-sales service and warranty records and structured household surveys. Measures will include household- and system-level factors such as stove functionality, repairs, service interactions, user-reported experiences, and contextual factors influencing adoption and use among both purchasing and non-purchasing households.

Reach will be defined as the extent to which households are exposed to implementation strategies. This will be measured as the proportion and characteristics of participating households reporting exposure to at least one implementation strategy, complemented by the number of individuals reached in each village based on program records of participation or contact by activity type. Results will be summarized across households and villages and stratified by household poverty status

Organizational adoption will be defined as the uptake and delivery of assigned implementation strategies by implementing organizations and delivery actors, including initiation and extent of delivery across planned components. Data will be collected through a combination of program implementation records documenting organizational participation and strategy delivery, and key informant interviews with representatives from each organization. Qualitative and quantitative data will be analyzed together to generate a comprehensive view of organizational adoption.

Implementation fidelity will be defined as the extent to which core components and intended sequencing of implementation strategies are delivered as planned, including systematic documentation of adaptations (type, timing, and rationale), distinguishing planned adaptations from reactive modifications. Data will be collected through a combination of structured implementer surveys, routine process evaluation tools, and key informant interviews with delivery actors. Qualitative and quantitative data will be analyzed together to generate a comprehensive view of implementation fidelity.

Implementation dose will be defined as the quantity and quality of delivery of implementation strategies, including frequency, number, and duration of contacts (dose quantity), as well as the scope of content delivered and level of engagement (dose quality). Data will be collected through a combination of structured implementer surveys, routine process evaluation tools, and key informant interviews with delivery actors. Qualitative and quantitative data will be analyzed together to generate a comprehensive view of the implementation dose.

Implementation acceptability will be defined as the perceived satisfaction, appropriateness, and feasibility of implementation strategies among implementers and participating households. Data will be collected through a combination of program records, key informant interviews with delivery actors, and structured household surveys. Qualitative and quantitative data will be analyzed together to generate a comprehensive view of implementation acceptability.