SOFTWARE QUALITY ASSESSMENT USING EXPERT ASSESSMENT

Abstract

The article presents a method of calculating the quality assessment of the software tool Visual Studio Code by an expert method is proposed. This method is based on the calculation of experts' evaluations according to a certain set of criteria that most fully depict the quality of the software according to various parameters. To evaluate the quality of the software, ten quality indicators for the software were selected and analyzed, which most comprehensively characterize the software in terms of adaptability to the field of application and further improvement at the request of the client. 8 experts took part in the evaluation of this software. After conducting a survey of experts, scores were obtained for each quality indicator and the average value of software quality was calculated. The number of users of the software may change over time, and so will user ratings. Since experts from this group can change and their assessment also changes, it was decided to call experts from this group dynamic experts. The developed method takes into account assessments from experts from a dynamic group at a certain point in time, for example, at the time of the release of the first version of the developed software and at the time of the release of the next version of the software. Between these two time stamps, end users can change their attitude towards software quality. The methods of obtaining expert assessments of the values of quality indicators for the purpose of methodological unity should be based on the general method of quality assessment, the rules for choosing the nomenclature of quality indicators and current methodical materials for the assessment of a given quality indicator of the evaluated software. Experts are invited to give a rank assessment of a predetermined number of product quality indicators. Rank assessment is reduced to indicating the degree of importance of each indicator by rank. The agreement of experts' opinions regarding the importance of each property is evaluated as the coefficient of variation of experts' opinions for each and every quality indicator. The larger the value of the coefficient, the lower the consistency of experts' opinions regarding the importance of the i-th indicator. If the ratio is <10%, the agreement of experts' opinions is considered high, if the ratio is <15%, it is above average, if the ratio is <35%, it is below average, and if the ratio is >35%, it is considered low. The concordance coefficient is determined to assess the overall consistency of experts' opinions. The agreement of experts' opinions will be better, the closer the concordance coefficient (W) is to one. The value W=0 indicates complete indifference or inconsistency of experts' opinions. The last step is to evaluate the value of each single quality indicator, taking into account its significance coefficient and the overall average value of software quality.