On adjusting productivity measures for multiple authorship