A macro analysis of productivity differences across fields: Challenges in the measurement of scientific publishing

While many studies have compared research productivity across scientific fields, they have mostly focused on the hard sciences, in many cases due to limited publication data for the softer disciplines; these studies have also typically been based on a small sample of researchers. In this study we use complete publication data for all researchers employed at Norwegian universities over a 4-year period, linked to biographic data for each researcher. Using this detailed and complete data set, we compare research productivity between five main scientific domains (and subfields within them), across academic positions, and in terms of age and gender. The study's key finding is that researchers from medicine, natural sciences, and technology are most productive when whole counts of publications are used, while researchers from the humanities and social sciences are most productive when article counts are fractionalized according to the total number of authors. The strong differences between these fields in publishing forms and patterns of coauthorship raise questions as to whether publication indicators can justifiably be used for comparison of productivity across scientific disciplines.