A Revised Procedure for the Calibration of the Activity-Wheel

The reliability of Shirley's method of calibration (1928)1 has been approved by Munn,2 by Rundquist and Bellis,3 and by Zieve.4 We hope to show that, in spite of use and currency of this method, Shirley failed to support her conclusion that revolving-drum technique is a reliable means of measuring spontaneous activity of animals. Shirley realized that a reliability coefficient, however high, is meaningless unless measuring instrument is itself consistent; in this case, unless all cages involved in measurement are equal in 'difficulty.' She says of her own study that the variability within a normal group of animals was found to be so great that a scrutiny of consistency of revolving-cage method of measuring activity was imperative. .. . It is obvious that, unless cages themselves are practically equivalent, any differences observed in performance of different animals may indicate nothing but a variation in characteristics of cages. This equivalence Shirley fails to demonstrate. The fundamental error lies in her calibration. Note her method.