Key points are not available for this paper at this time.
Data interval bias, the biased estimation of advertising carryover in aggregate data, can be viewed as a misinterpretation of the aggregate advertising-sales relationship due to missing micro advertising data. This paper argues that if the researcher does not explicitly model the firm's advertising decisions, he will incorrectly interpolate the missing data and thereby allow the firm's decision behavior to influence inferences about advertising carryover. Drawing upon a general model of advertising decision behavior, the expected aggregate form of the Koyck relationship is developed and compared to existing bias correction methodologies. Although it is difficult to find any parsimonious procedure which is robust with respect to all types of decision behavior, allowing lagged advertising to enter the classical Koyck equation emerges as a simple method of obtaining a reasonable estimate of advertising carryover in aggregate data.
Gary J. Russell (Mon,) studied this question.