Are Null Results Becoming an Endangered Species in Marketing?

Abstract

Editorial procedures in the social and biomedical sciences are said to promote studies that falsely reject the null hypothesis. This problem may also exist in major marketing journals. Of 692 papers using statistical significance tests sampled from the Journal of Marketing, Journal of Marketing Research, and Journal of Consumer Research between 1974 and 1989, only 7.8% failed to reject the null hypothesis. The percentage of null results declined by one-half from the 1970s to the 1980s. The JM and the JMR registered marked decreases. The small percentage of insignificant results could not be explained as being due to inadequate statistical power. Various scholars have claimed that editorial policies in the social and medical sciences are biased against studies reporting null results, and thus encourage the proliferation of Type 1 errors (erroneous rejection of the null hypothesis). Greenwald (1975, p. 15) maintains that Type I publication errors are underestimated to the extent that they are: ... frightening, even calling into question the scientific basis for much published literature. Our paper examines the publication frequency of null results in marketing. First, we discuss how editorial policies might foster an atmosphere receptive to Type I error proliferation. Second, we review the evidence on the publication of null results in the social and biomedical sciences. Third, we report on an empirical investigation of the publication frequency of null results in the marketing literature. Fourth, we examine power levels for statistically insignificant findings in marketing to see if they are underpowered and thus less deserving of publication. Finally, we provide suggestions to facilitate the publication of null results.

Keywords: File drawer problem, null results, publication bias, statistical power analysis