This paper deals with measurement of ground water arsenic contamination. The focus is on using a proper index for the severity of contamination, rather than just using the proportion of observations above a threshold level. We specifically focus on the Contamination Severity Index (CSI) proposed in Sen (2016, Sankhya). An alternative estimator in contrast to the one given in Sen (2016)  is used here that is useful for small number of observations. The data used is that collected by British Geological Society(BGS) and the BD Department of Public Health Engineering (DPHE) during 1997-2001. Their analysis was based on the simple proportion of the observations above a threshold level, where as the CSI measure adequately takes into account the severity of the observations also. We have also segmented the data into three categories of wells according to the depth of the wells instead of just the two categories, namely `deep' and `shallow' wells. It is emphasized in this manuscript that the comparison of areas with average arsenic (As) level  to determine As severity is not appropriate as the regression of CSI on AAs is highly nonlinear and seemingly non-heteroscedastic; where as the CSI  index proposed in Sen (2016), shows a clear picture, especially when the values are adjusted according to average log depth of the wells sampled at the thana and district levels.