Dependence of publication-rate on size of some university groups and departments in U. K. and Greece in comparison with N. C. I., USA

Abstract

In a series of studies aimed at investigating the dependence of per-capita research output (R), of an interacting group of research workers, on the size of the group, the author had shown that the per-capita research output of various research groups and institutes in U. S. A., U. K., Pakistan and Bangladesh shows an initial approximately linear rise, followed by one or more mixima, the first one being at group size of 6 to 8 persons. In the present communication, we present a fine analysis of the reported data for (a) physics departments of U. K. universities (in 1985–86) and (b) mathematics departments of two universities in Greece (from 1975 to 1984), using close sampling-intervals of ΔN=2 and 3 for group-sizes. The results of this reanalysis show that the data for U. K. physics departments exhibits a series of peaks ofper-capita research output (R) at N=11, 19, 25, 36, 46, etc., which compare well with the corresponding maxima already found in the 1977per-capita output of National Cancer Institute, U. S. A., at N=7, 15, 26, 34 and 44. Comparison of these two yields the followingmean positions for the five peaks viz N=9±2, 17±2, 26±0, 35±1 and 45±1. These appear to be close to multiples of 8.5, indicating the possibility that a sub-group of 8 to 9 persons could be forming abasic unit of interaction in these particular research groups. The data from the mathematics departments of two Greek universities, which falls in the range of N=20 to N=44, also shows two maxima, of per-capita output at N=27 and 34.5 (and possibly one at about 18), which fit in well with the pattern described above. It appears likely that the above concept could open up new avenues in management practices. Accordingly, further studies are in hand on the relevant characteristics of the output of various institutes and, if possible, a fuller study of size and nature of the sub-groups noted above.