Calculation
Four counting methods are used for authors; university rankings use the first two. The definitions follow Chua, Cao, Cousins and Straub (JAIS, 2002), the standard counting methods in research productivity studies of the IS field.
Normal
Every co-author of a paper receives 1. A university receives at most 1 per paper, however many of its authors are on it.
Adjusted
Each of the n authors receives 1/n. A university with p of the n authors receives p/n.
Weighted
Credit steps down with team size: 1 for a single-authored paper, 0.7 for two authors, 0.5 for three, 0.3 for four or more.
Straight
Only the first author receives credit.
The paper-type filter is ours: conference proceedings distinguish full research papers from short papers (research in progress), and the two tell different stories. AMCIS is roughly 40% short papers, so a normal count rewards volume that a full-paper count does not.
Movement arrows
The small arrow beside each rank shows how that position changed once the most recent year in view is counted. It compares two rankings: the one you are looking at, and the same ranking with the last selected year removed. Selecting 2021 to 2026 therefore compares against 2021 to 2025, and ▲5 means the author or university stands five places higher than they did before the 2026 papers were added.
The comparison is recomputed from your current filters, not read from a stored table, so the arrows stay correct when you change conference, region, paper type or counting method. A new marker means no papers at all in the earlier window, which is a first appearance rather than a large rise. Selecting a single year leaves nothing to compare against, so the arrows are hidden.
Two things are worth keeping in mind when reading them. Movement in a rolling window reflects papers leaving the window as well as papers being added, and near the middle of a long ranking a single paper can move someone many places, because thousands of people sit within a paper or two of each other. Large movements at the top of the table are meaningful, large movements at rank 4,000 usually are not.