Kritzinger, PierreHuman, Schalk WilliamChakraborti, Subhabrata2015-11-052015-11-052014-07P. Kritzinger, S. W. Human & S. Chakraborti (2014) Improved Shewhart-type Runs-rules Nonparametric Sign Charts, Communications in Statistics - Theory and Methods, 43:22, 4723-4748, DOI: 10.1080/03610926.2012.729637.0361-0926 (print)1532-415X (online)10.1080/03610926.2012.729637http://hdl.handle.net/2263/50348Runs-rules are typically incorporated in control charts to increase their sensitivity to detect small process shifts. However, a drawback of this approach is that runs-rules charts are unable to detect large shifts quickly. In this paper improved runs-rules are introduced to the nonparametric sign chart, to address this limitation. Improved runs-rules are incorporated to maintain sensitivity to small process shifts, while having the added ability to detect large shifts in the process more efficiently. Performance comparisons between sign charts with runsrules and sign charts with improved runs-rules illustrate that the improved runs-rules are superior in performance for large shifts in the process, while maintaining the same sensitivity in the detection of small shifts.en© Taylor & Francis Group, LLC. This is an electronic version of an article published in Communications in Statistics Theory and Methods, vol. 43, no.22, pp. 4723-4748, 2014. doi : 10.1080/03610926.2012.729637. Communications in Statistics Theory and Methods is available online at : http://www.tandfonline.comloi/lsta20.NonparametricSign chartShewhartControl chartRuns-rulesImproved runs-rulesAverage runlength (ARL)Control limitStatistical process control (SPC)Postprint Article