Secondary Logo

Journal Logo

Statistical Clarification

Fisher, Dennis

doi: 10.1097/00000539-200104000-00051
Letters To The Editor: Letters & Announcements
Free
SDC

Durect Corporation, Cupertino, CA

To the Editor:

Two readers (1) recently criticized authors for a statistical error. The authors (2) had reported that a range of 33 values from 10 to 60 had a median value of 10. The readers wrote “at least one of these values must be in error, as it is not possible to have the minimum and the median values identical”(1). Although the authors acknowledged that the data were reported incorrectly (3), the readers are themselves incorrect–the median and the minimum CAN be the same. If the set of 33 values contained 17 or more values of 10, the median and the minimum would both be 10. Although this is not instinctive (the readers assume that the median must lie somewhere BETWEEN the minimum and the maximum), it is true.

Dennis Fisher

Back to Top | Article Outline

References

1. Armstrong S, Russell WJ. The onset of MH [letter]. Anesth Analg 2000; 91: 1560.
2. Allen GC, Brubaker CL. Human malignant hyperthermia associated with desflurane anesthesia. Anesth Analg 1998; 86: 1328–31.
3. Allen GC. In response to “The onset of MH” [letter]. Anesth Analg 2000; 91: 1560.
© 2001 International Anesthesia Research Society