SUMMARY
The article discusses the necessity of creating competency based tasks for computer educational models. This requires clarification of competency concept as well as the methods of its measurement. Several approaches to understanding competency and its definition have been analyzed. It is obvious that the majority of those do not describe development of measurement instruments for competency levels. Some of the discussed approaches make it possible to split the competency into separate components and consider the activities leading to upgraded competencies. The author provides definitions of competency and competency level. Competency is a mental construct which is a certain unity of a person’s skills required within definite context of activity for a definite set of possible tasks. Competency always describes “unity of two” – 1) set of tasks in the context; 2) the performer. Competency level is an index, characterizing the probability of successful performing of the task (on the given level) out of a definite set (variety) of tasks in the defined context by the definite performer. The experimental and indirect methods of competency measurement have been described. It is possible to view the construction of the competency from two different points: which parts (ingredients) and which components it includes. Parts (ingredients) are knowledge, skills, behaviours, attitudes, etc. Components are minor competencies aimed to solve minor auxiliary tasks. Thus, in order to solve a competency task it is necessary to solve a few easier tasks which would complete some zones of competency. The proposed approach describes the theory of competency level and structure measurement being a basis for competencies forming management. The mentioned approach requires more detailed development which is viewed as the matter of further research and publications.