Keep Missing Measurements Separate from Zero - Practice 1

5 questions0m 0s

Learn the skill

Missing is not zero. A mean calculated only from observed locations describes that observed subset and can be biased for the whole region if missingness is selective. These fictional records give equally weighted observed values.

Worked example

Two observed values 10 and 20, plus one missing value, give an observed-only mean of 15, not 10. The full-region mean is unknown.

Question 1 Fictional locations 1: values 10, 20, and missing. What is the mean of the two observed values only? Answer

Question 2 Fictional locations 2: values 12, 22, and missing. What is the mean of the two observed values only? Answer

Question 3 Fictional locations 3: values 14, 24, and missing. What is the mean of the two observed values only? Answer

Question 4 Fictional locations 4: values 16, 26, and missing. What is the mean of the two observed values only? Answer

Question 5 Fictional locations 5: values 18, 28, and missing. What is the mean of the two observed values only? Answer