How to analyze publication time trends by correspondence factor analysis: Analysis of publications by 48 countries in 18 disciplines over 12 years

Although different countries laid emphasis on publication in different disciplines, the overall tendency was toward greater uniformity in publication patterns with time, which was not a linear but an elastic scale.

Abstract

Abstract This study is a follow‐up to a published Correspondence Factorial Analysis (CFA) of a dataset of over 6 million bibliometric entries (Doré et al. JASIS, 47(8), 588– 602,1996), which compared the publication output patterns of 48 countries in 18 disciplines over a 12‐year period (1981–1992). It analyzes by methods suitable for investigating short time series how these output patterns evolved over the 12‐year span. Three types of approach are described: (1) the chi 2 distances of the publication output patterns from the center of gravity of the multidimensional system—which represents an average world pattern—were calculated for each country and for each year. We noted whether the patterns moved toward or away from the center with time; (2) individual annual output patterns were introduced as supplementary variables into an existing global overview covering the whole time‐span [CFA map of (countries × disciplines)]. We observed how these patterns moved about within the map year by year; (3) the matrix (disciplines × time) was analyzed by CFA to derive time trends for each country. CFA revealed the “inner clocks” governing publication trends. The time scale that best fitted the data was not a linear but an elastic scale. Although different countries laid emphasis on publication in different disciplines, the overall tendency was toward greater uniformity in publication patterns with time.