Course · Intermediate · Indicators and Measurement
Indicator Design and the LogFrame
Write indicator definitions precise enough that two analysts computing them independently get the same number, and trace each one back to the change it is supposed to evidence.
What you will be able to do
- Trace an indicator back to the change it is meant to evidence, and say what the LogFrame lost on the way from the Theory of Change
- Write an indicator reference sheet complete enough that a second analyst reproduces your number without asking you a question
- Choose a denominator you can defend in a review, and state what your choice excludes
- Distinguish reach, coverage and cumulative counts, and stop a monthly series from counting the same person twelve times
- Adopt a standard indicator definition where one exists, and say precisely where yours departs from it
- Set a baseline and a target that survive a mid-term review, and name the behaviour each one will encourage
Standards and methodologies
This audience is judged on defending a number in a review, not on producing it. Module 2 got the data to a state where a number can be computed at all. This course is about which number, and why, and how to write it down so that the person who replaces you computes the same one.
The central artefact is the indicator reference sheet: a page per indicator carrying the numerator, the denominator, the disaggregation, the frequency, the source and — the field almost every template omits — the decision the number informs. An indicator that informs no decision is a reporting cost, and naming the decision is the fastest way to find out that you have several.
The recurring failure this course exists to prevent is the one the Data Quality Assessment course kept finding underneath everything else: two people computing different things under one indicator name. That is not a data quality problem and no amount of cleaning fixes it. It is a definition problem, and definitions are cheap to fix and permanent once fixed.
Both Python and R throughout, though this course has more prose and fewer pipelines than the rest of the module — the hard part of an indicator is rarely the arithmetic. Worked examples run against the DHIS2-shaped vaccination extract, where the penta1-to-penta3 dropout is 13.8% and every coverage figure rests on a denominator somebody projected from a 2015 census, and against the MUAC screening register, where 4,218 screenings are not 4,206 registrations and neither is the number of children.
You should have done Data Analysis Foundations. Module 2 is not a prerequisite, but every example assumes you already know why a missing report is not a zero.
Practice
Reading the lessons is not the same as having done the work. Each of these applies the course to a dataset it did not teach on.
- Lab · 120 minThree ladders, three denominatorsWrite reference sheets for the JMP water, sanitation and hygiene service ladders, compute all three from one household survey, and discover that the three indicators cannot share a denominator.
- Exercise · 45 minOne numerator, four reportsFour organisations compute the same indicator from the same register and publish four different numbers. Nobody made an error. Reproduce all four, then write the reference sheet that would have prevented the meeting.
Progress
Take it offline
The whole course as a typeset PDF — every lesson, every code example, the data dictionary and the indicator definitions. Generated from the same source as this page.
The LaTeX source ships alongside each PDF, so an organisation can rebrand the handout or fold a lesson into its own training pack.