Information and misinformation in bibliometric time-trend analysis

A diachronous time-series of bibliometric data (using all data available) suggests rising normalised citation impact (nci) for Germany and other G7 nations, while China suffers a decline in later years of any series. This is shown to be a consequence of the time-series, which has led to an erroneous interpretation of real trajectories. A synchronous series (using fixed time windows) based on the final year suggests a lower trajectory while a diachronous series tracking the fate of a single publication year reveals that nci progressively falls for Germany and the USA whereas it climbs for China. This has implications for research policy and for the interpretation of changes in the competitive research environment in the presence of dynamic growth. By analogy, this may extend to institutional as well as national comparisons. It has implications for analytical methodology, supporting prior suggestions that recent papers should be omitted from citation analysis.