Kliniki Yangu · Proposed formative study
Observe the task.
Listen to the learner.
A 23-minute protocol for a future review. No participant results have been collected.
Before inviting participants
Have a Kenyan maternal-health professional check the context and a fluent Kiswahili reviewer check every instruction, response and translation. Recruit consenting adults through an appropriate partner. Confirm that participation is voluntary, has no effect on care and can stop at any point. No clinical history, real messages or contact information is required.
Use an anonymous test code. Record consent separately if required by the partner. Do not record audio or video without separate consent. Ask about language and phone experience directly; do not infer ability from pregnancy, income, education or location.
1. Establish the context · 3 minutes
- Which language would you like to use?
- What do you usually use a phone for?
- What do you do when a message is difficult to understand?
- Would you prefer to try alone or read together?
2. Baseline task · 3 minutes
Show a separate fictional message: “Dear Neema, your clinic appointment TMH-APP-261002-0034 is scheduled for 20 Oct 26 at 11:00 at Tumaini Hospital. Please bring your clinic card.” Ask the participant to identify the date, time and clinic, then explain what they would do if the time were unclear. Do not use the lesson's assessment message as the baseline.
For Kiswahili: “Mpendwa Neema, miadi yako ya kliniki TMH-APP-261002-0034 imepangwa tarehe 20 Okt 26 saa 11:00 katika Hospitali ya Tumaini. Tafadhali leta kadi yako ya kliniki.” Apply the same task and assistance rules. Rotate equivalent baseline and follow-up examples across participants where practical.
3. Try the lesson · 10 minutes
Say: “This is a practice lesson with made-up appointments. Please try it at your own pace. We are checking the design. You can ask for help or stop whenever you wish.” Observe before offering help. Record each prompt you give. Read aloud if requested, distinguishing access support from hints that reveal an answer. Observe whether inline phrases are easy to discover and tap, including on small screens. Ask whether “visit started” and “consultation completed” give a future appointment. Check that numbers inside an appointment ID are not treated as the next appointment date.
4. Transfer and discussion · 5 minutes
Use the lesson's final message for task transfer. Ask: “What will you do if the message is unclear?” Then: “Which part was difficult?” and “What would make this easier on your own phone?” Ask the participant to explain one instruction in their own words.
5. Close · 2 minutes
Thank the participant. Reconfirm that the examples are fictional. Offer the blank card as an optional learning aid; their clinic supplies real appointment details. Do not interpret medical questions: direct them to their care team.
Observation sheet
| Record | Observation |
|---|---|
| Anonymous code, language, device | ________________________________ |
| Baseline: date / time / clinic correct | ________________________________ |
| Can open the message without help? | ________________________________ |
| Recognises completed status and missing next date? | ________________________________ |
| Where is help requested? | ________________________________ |
| Final: date / time / clinic correct | ________________________________ |
| Hints used / access support provided | ________________________________ |
| Can identify an appropriate help route? | ________________________________ |
| Participant explanation or feedback | ________________________________ |
| Design change to try next | ________________________________ |
Report counts with denominators and distinguish independent success, success with access support and success after an answer hint. Before/after differences in a small formative study cannot establish causality or improved antenatal attendance. Test translation and voice availability on the devices people actually use.
Kevin Gakarau · Independent portfolio prototype · September 2026