We report on how different accounting procedures affected the counting of citations of physics papers,thereby affecting the rankings of research impact at country level.Using 1989-2008 data,we also report the counting inflation ratio between different accounting procedures.We found that the normal count (all author count) procedure yielded a rather different ranking from those of straight count and adjusted count approaches.It also tended to favor the European countries,while the newly developed countries received more credits in the adjusted and straight count procedures.