A non-parametric statistical analysis in the measurement of outdoor gamma exposure to the residents around Trombay.

ABSTRACT

During this study, non-parametric statistical methods were used to validate the measured gamma dose rate with the calculated one, around Trombay. Portable dose rate digital gamma spectrometry system (target fieldSPEC) was used for in situ measurement of external gamma (gamma) dose rate (measured) with the range of 1 nSv/h-10 Sv/h. The activity concentration of U-238, Th-232, K-40 and Cs-137 in the soil and their respective external dose-conversion factor (nSv/h/Bq/kg) was used to evaluate the gamma dose rate (calculated). Non-parametric statistical tool like Box- and -Whisker Plot, Spearman's (rho) rank Correlation coefficient, the Wilcoxon/Mann-Whitney test and chi(2) distribution test have been applied for validation. The randomness or discrete behaviour of measured and calculated dose rate was obvious from the Box- and -Whisker Plot as mean and median of the two are not equal. The inter quartile range (Q3-Q1), which explains about the dispersion of measured and calculated dose rate were also evaluated and found to be 10 and 16 microSv/y, respectively. The linear association between the order of ranks of the two dose rates was established by using Spearman's (rho) rank correlation that showed a coefficient of R = +0.90 with the intercept +1.9, whereas Pearson's correlation was observed with a coefficient of R = +0.93 with the intercept -25.6. Wilcoxon/Mann-Whitney test shows that, medians of the calculated and the measured dose rate as significantly different under the assumption of null hypothesis and measured dose rate was made to the normal distribution by applying Z-statistics. Value of chi(2) was calculated and found to be 284.95, which was very much greater than the critical value of chi(2)(0.05) = 43.77 at a degree of freedom 30, concluding that there is a highly significant difference between the measured and calculated dose rate at 5% significance level.