Part 4 · How do you design a study?
Module 7 · How do you plan a CNA study?
Three decisions before you touch the data
- Define a clear, measurable outcome with a threshold.
- Draw up a disciplined list of candidate conditions.
- Write down a causal theory before analysing.
A CNA study is mostly won or lost at the planning table, long before R enters the picture. Three decisions matter most.
Define the outcome
State exactly what you are trying to explain, and how you'll recognise it when you see it. For example: "Low routine immunization coverage" means DTP3 below 50%. The threshold is part of the definition, not an afterthought.
Identify plausible conditions
List the factors that program experience says could matter: no trained staff, broken refrigerator, no supervision, no community mobilisation, no outreach funding, thin staffing. Keep the list disciplined. Every condition should have a real reason to be there, not just a vibe.
Write down a causal theory
Before you analyse anything, commit to what you expect. For example: low coverage may come from a broken cold chain plus no supervision, OR from no community engagement plus no outreach. Writing it down ahead of time keeps you honest when the results land.
Sketch a quick study for a problem you care about.
- What is your outcome, and exactly where is the cut-off between high and low?
- Which five or six conditions genuinely belong on your list, and why each one?