Periodic dose measurements are conducted using the thermoluminescent method, using TLD along with Harshaw 4500 and Harshaw 6600 readers. For accurate dose assessment of employees, it is essential to calibrate the dosemeters or determine their correction coefficients. We have done it using the internal source of the Harshaw 6600 (Sr-90). An important factor in assessing uncertainties in thermoluminescent measurements is the uncertainty arising from the element correction coefficients (ecc) values, therefore it is necessary to use advanced techniques to study the distribution of values ecc. Approximately 1000 dosemeters were analysed using statistical methods. The distribution of ecc coefficients was optimized and fitted to well-known distributions. A null hypothesis regarding acceptable fit was tested using the chi square - test. The results showed that ecc coefficients do not follow a known statistical distribution due to variations in work periods and the number of measurements performed. The same analytical approach was applied to 150 newly calibrated dosemeters. In this case, it was determined that their distribution could be fitted to a known statistical distribution. Estimating measurement uncertainties by studying the distributions of values in depth is more accurate than using simplified methods.
Gega et al. (Wed,) studied this question.