A confirmatory analysis of a previously reported structure for the Level of
Expressed Emotion (LEE) scale was conducted with data from 75 volunteers.
Three principal components were extracted and rotated to maximum congruence
with a target based on previous research. The fit was found to be highly s
ignificant and retest correlations for the subscales of the LEE were good.
However, the revised LEE scale could probably be improved, especially by th
e addition of a subscale designed to measure perceived criticism.