Anatolii Maslak

The relevance of this study is explained by the fact that the quality of school education is determined by the structure of teaching staff to a considerable extent. The following factors are investigated: teachers’ category (without category, first category, higher category), teachers’ age, work experience and job place (rural or urban). The response is the percentage of teachers working in each segment, determined by a combination of the factors levels. The multifactorial analysis of variance revealed the significance of factors and their interactions. The study involved all schools of Krymsk district, Krasnodar region, 1100 teachers in all.

There are investigated factors influencing the degree of test variants parallelism– the range of variation and the number of test tasks in a bank of tasks, the range of examinees’ proficiency and the number of test variants. The research was conducted within the framework of latent variables theory. The level of examinees’ proficiency and difficulty of test tasks are estimated in logits. The complex factorial experiment was fulfilled on the base of simulation. For statistical processing of the simulation experiment results the four-factorial analysis of variance was used. It is shown that degree of test variants parallelism mostly depends on the variation range of test tasks in the bank, then on the test length and the number of test variants. The number of test tasks in the bank does not influence parallelism of the test variants. Two two-factorial interactions are significant: the influence effect of test tasks variation range in the bank depends on the number of test tasks in the test and the influence effect of test tasks variation range in the bank depends also on number of the test variants.

The estimation of efficiency of formation creativity program for students is fulfilled. Research was con-ducted within the framework of the theory of measurement of latent variables, based on Rasch model. The latent variable «creative self-efficacy» is defined the based on set of indicators. Statistical analysis has shown that all indicators are compatible with each other and consequently all set of indicators can be used as the measuring tool. The questionnaire possesses good differentiating ability and corresponds to estimation of creative self-efficacy of students. Measurement on a linear scale of the latent variable «creative self-efficacy» has allowed to carry out the multifactorial analysis of variance students’ creative self-efficacy depending on their specialty and grade. Results of the analysis of variance have shown, that based on creative self-efficacy there are significant distinctions between students depending on their specialty and a course.

The paper applies the theory of latent variables for measurement of pupils’ patriotism in comprehensive school. It was found out that the level of pupils’ patriotism varies in a small range. The authors have not revealed statistically significant distinctions between pupils’ patriotism in the classes taking part in the survey. The levels of patriotism of girls and boys are practically the same.

Within the framework of the theory of latent variables, the latent variable “physical aggression of teenagers” is measured on a linear scale. It is shown that the used set of indicators (questionnaire points) allows measuring both low, and high level of teenagers’ physical aggression. The analysis of teenagers’ physical aggression depending on grade and gender of teenagers is carried out. The analysis of variance has shown that on the average young men are more aggressive, than girls are. However, there is no statistically significant difference on an investigated latent variable between classes.