Abstract
This paper examines the properties of the Item Efficiency Index proposed by Neill and Jackson (1976) for minimum redundancy item analysis. Regarding the components of the index it is suggested that the item score be deducted from the scale scores in which it is included for the calculations of the item-scale correlations and that in general the point-biserial correlations be used instead of the biserial ones. The addition of a parameter is proposed and problems regarding the estimation of the parameters are pointed out. Finally, several limitations in the interpretation and applicability of the index are discussed.

This publication has 3 references indexed in Scilit: