devericklasco/surveycto-coffee-farmers-uganda

# surveycto-coffee-farmers-uganda This repository is a public SurveyCTO/XLSForm work sample for field operations context: an AI advisory tool for smallholder coffee farmers in Uganda. The package is designed to demonstrate strong instrument engineering and field readiness, not agronomy depth. Created by Erick Busuulwa (devericklasco). ## How this maps to Field Coordinator responsibilities - Instrument design: Structured modules from setup through closeout, with practical skip logic and validation. - Testing and debugging: Built-in constraints, safe calculations, and explicit testing notes for core paths. - Translation support: Parallel English and Luganda labels/hints to support multilingual deployment. - Enumerator readiness: A concise enumerator guide with consent script, probing standards, and GPS steps. - Data quality assurance: Metadata capture, constraint messages, calculated indicators, and outcome coding. ## XLSForm technical features included - Module-based design using `begin_group` and `end_group`. - Skip logic using `relevant` expressions across consent, fertilizer, sales, and incomplete interview branches. - Constraints with bilingual error messages (age, land, coffee acreage, yield, price, phone number, IDs). - Safe calculations: - `yield_per_acre_kg = coffee_yield_kg / coffee_acreage` with divide-by-zero protection. - `duration_sec` from `start` and `end` using decimal datetime math. - `gps_accuracy_m` parsed from geopoint accuracy. - Metadata and field ops variables: `start`, `end`, `today`, `deviceid`, `enumerator_id`, GPS, duration, and form status. - “Other (specify)” patterns in multiple modules (income source, fertilizer type, challenges, advice topics, channels, incomplete reason). ## Import and testing in SurveyCTO (high level) 1. Open SurveyCTO server and create a new form. 2. Upload `coffee_farmer_survey.xlsx`. 3. Run the form tester to verify: - consent refusal path, - required field behavior, - skip logic, - constraints and message …