1. M.M. Qurashi, The optimum size of research groups for maximum effectiveness. Part I- Statistical formulation and analysis of the data for U.K., Canada and Pakistan,Pakistan J. Sci. Ind. Res., 12 (1969) 1–7. This shows the presence of two maxima, one at N=28 and the other at N=70 in the size distributions, of applied research groups.
2. Ibid.,Pakistan J. Sci. Ind. Res., 12 (1969–70) 315–323. This provides a theoretical explanation for the empirical size distributions, using three assumptions viz. (i) the maxima correspond to max. percapita output, i.e. efficiency, (ii) per-capita output of a group of N researchers is proportional to (N)1+m, where 03. Ibid.,Proc. Pakistan Acad. Sci., 9 (1972) 109–117. This provides direct evidence for the variation of per-capita output of groups from zero to maxima at N ∼20 and N∼80, followed by a slow fall towards zero with increasing N.
4. Ibid.,Proc. Pakistan Acad. Sci., 12 (1975) 1–13. This provides data to show that research groups in university department (and similar units) also show two maxima in the size distributions, but at N=5±1 and N=15±5.
5. M.M. Qurashi, A study of procedures of selecting and channelizing scientists for research and development. Part III-Cyclic variations in output,Pakistan J. Sci. Ind. Res., 12 (1969) 499–504. In the course of a study of the annual per-capita output of several established Pakistan Scientists, a cyclic variation is found: two prolific scientists' groups are also studied, data on one of which (M.H.K.'s group) are now utilized in the present paper.