Abstract
This article investigates the assumption that extreme ratings on the semantic differential are preferable and that common agreement exists among users in the interpretation of items. These assumptions were investigated by measuring ideal ratings for three source types.

This publication has 4 references indexed in Scilit: