How this calculator works
Count observations below the score and equal to it. Strict ranking uses only those below; inclusive ranking adds all equals; midpoint ranking adds half the equals. Average percentage rank uses the mean one-based occupied rank when matches exist, and the strict proportion when there are no matches. These rules answer slightly different questions, especially in small lists with many ties.
Formula
With B below, E equal and N total: strict = B/N; inclusive = (B + E)/N; midpoint = (B + E/2)/N; average rank = (B + (E + 1)/2)/N when E > 0.
Worked example
For 1, 2, 3, 3, 4 and score 3, B = 2 and E = 2. Midpoint rank is 60%, inclusive is 80%, strict is 40% and average rank is 70%.
Advertisement
Frequently asked questions
Must the score occur in the data?
No. With no exact match, all four conventions give the fraction of observations below the score.
Which method is best?
Use the rule specified by the test, data provider or analysis. Report it when ties could affect interpretation.
Does a 90th-percentile rank imply exactly 10% scored higher?
That depends on the tie convention. Inclusive ranking has that direct interpretation; midpoint and average-rank conventions allocate ties differently.