Evaluating the performance of photovoltaic modules with high precision during field testing is difficult since performance parameters are influenced by several factors, including some which are normally not measured. This work shows that data filtering based on four basic measurement parameters can improve the precision markedly by restricting the analysis to well-defined measurement conditions. Field measurement data from CIGS photovoltaic modules were filtered and analyzed, and a methodology for selecting the best filtering criteria was developed. A comparison with a traditional method shows that the variance in the data can be reduced by as much as 70-80% with suitable filtering conditions. The same filtering reduces the amount of data points by 50%. The method also includes a calculation of temperature coefficients which takes into account time-dependent changes in performance parameters. The results presented in this paper can be used as a tool for planning field tests of photovoltaic modules, for versatile analyses of measured data and for the detection of changes in performance parameters at an early stage of field testing.