Differences in productivity and impact across the different computer science subareas

Can we quantitatively compare different computer scientist? There is a widespread belief among computer science researchers that different subareas within Computer Science (CS) have different publishing practices (production throughput per year, preference either for journals or conferences, number of citations, etc.) making the use of a unified evaluation criterion unfair. In this paper we present productivity measures (both journal and conference productivity) and impact measures (citations per paper and H-index) for a random set of researchers in 17 different CS subareas. This research quantifies the mentioned intuitions and pre-empirical impressions and shows that, indeed, there are different publication practices and impact value measures for the different CS subareas. However, we show that few of the differences are significative.