The impact of the science research council's policy of selectivity and concentration on average levels of research support: 1965–1974