MEPX
Chapter 8 of 10All chapters
  1. 1What is one row
  2. 2Rates and denominators
  3. 3Beyond the average
  4. 4Compared with what
  5. 5Aggregation traps
  6. 6Noise and significance
  7. 7Metrics and proxies
  8. 8Missing data
  9. 9Reading forecasts
  10. 10A reading routine

Chapter 8 of 10

Missing data

Absence is information.

Why it is missing

Missing at random is survivable; missing because of the value itself is not. Non-response from dissatisfied customers biases every satisfaction figure you compute.

  • Count and report missing values rather than dropping them silently.
  • Zero, blank and not applicable are three different things.

Filling gaps

Imputing a mean or carrying the last value forward invents data and understates uncertainty. Whatever choice is made should be stated with the results.