A Measurement Model to Assess Measure-Specific Factors in Multiple-Informant Research

In research employing multiple informants to assess the construct validity of properties of marketing organizations, both convergent and discriminant validity of the constructs may be overstated if measure-specific factors are not taken into account. The author presents a second-order confirmatory measurement model which explicitly models the organizational constructs of interest and measure-specific factors, as well as informant bias factors and random measurement errors. A comparative illustration is given and some implications for research on marketing organizations are discussed.