Specification of variables that reflect organizational processes can add an important dimension to the investigation of outcomes. However, many contextual variables are conceptualized at a macro unit of analysis and may not be amenable to direct measurement. In these situations, proxy measurement is obtained by treating organizational members as key informants who report about properties of the work group or organization. Potential sources of bias when using key informant methods in organizational survey research are discussed. Statistical procedures for assessment of rater-trait interaction as a type of informant bias are illustrated using data from a study in which multiple key informants were sampled to obtain proxy measurement of the organizational climate for caring among baccalaureate schools of nursing.