Unraveling Scientific Impact

The number of citations a paper receives is the most commonly used measure of scientific impact. In this paper, we study not only the number but also the type of citations that 659 marketing articles generated. We discern five citation types: application, affirmation, negation, review and perfunctory mention (i.e., citing an article only indirectly without really using it). Prior literature in scientometrics recognizes that the former three types, on average, signal a higher level of scientific indebtedness than the latter two types. In our sample, these three types of citation represent only 15% of all citations. We also find different determinants of citation behavior across citation types. Across the 49 determinants we included, only 13 have the same effect across all citation types, of which only 5 are statistically significant across all citation types. For instance, we find a significant inverted U-effect of challenging commonly held beliefs on citations counts, but only for three of the citation types: affirmation, review and perfunctory mention. Our results encourage scientific stakeholders to move beyond mere citation counts to assess a paper’s or a scholar’s scientific contribution, as well as to devote greater attention to the citation process itself.