Estimating firm-specific and relational properties in interorganizational relationships in marketing

Past research on interfirm relationships typically has relied on a single informant from one side of the relationship and has estimated the constructs of interest by summing the informant's answers to several measures posited as indicators of each construct. We demonstrate analytically that two systematic sources of measurement error, informant bias and measure specificity, have opposing effects in single-informant, sum-scale representations, and also provide empirical evidence of the relative magnitude of these confounds. Using an extremely rare dataset where there are multiple informant reports from each side of the relationships studied, significant perceptual agreement between partner firms is found for firm-specific and relational properties. The results provide empirical evidence on the magnitude and direction of informant biases and measurement specificity. We conclude with implications of the findings for substantive research on interorganizational relationships.