The iceberg hypothesis revisited

A study is described of the rank/JIF (Journal Impact Factor) distributions in the high-coverage Scopus database, using recent data and a three-year citation window. It includes a comparison with an older study of the Journal Citation Report categories and indicators, and a determination of the factors most influencing the distributions. While all the specific subject areas fit a negative logarithmic law fairly well, those with a greater External JIF have distributions with a more sharply defined peak and a longer tail—something like an iceberg. No S-shaped distributions, such as predicted by Egghe, were found. A strong correlation was observed between the knowledge export and import ratios. Finally, data from both Scopus and ISI were used to characterize the rank/JIF distributions by subject area.