Метод оценки сложности логических рассуждений

Abstract

The development of methods for assessing the various didactic characteristics of teaching texts and teacher messages is an urgent problem of modern teaching theory. Its solution allows to rank texts (descriptions, proofs, explanations, etc.) according to the degree of complexity, and this helps to optimize the learning process and increase the objectivity of assessing students' knowledge. In general, educational texts contain a descriptive-narrative and logical component; the share of the latter can be defined as the ratio of informative content of logical reasonings to the overall informative content of the text. The complexity (informativity) of the text can be found as a sum of the com-plexities of its constituent scientific terms and ordinary words. The semantic complexity of the term relative to the Z0 thesaurus is equal to the number of words that need to be spoken to explain the term to a student with the thesaurus Z0. It should be noted that each new mention of the term reduces its contribution to the overall complexity of the text. The proposed method for assessing the complexity of logical reasonings involves: 1) identifying all facts and logical rules used, the complexity of which exceeds the cho-sen level of knowledge; 2) expanding the text by adding all the statements necessary for conducting reasonings; 3) creation of text file text.txt containing verbal part of educa-tional text and verbally encoded formulas and figures; 4) creation of a dictionary (slo-var.txt file) containing the scientific terms used with indication of their level of com-plexity; 5) analysis of the text.txt file using the special computer program addressing to the dictionary which allows to determine the overall semantic complexity of the eva-luated text; 6) isolating logical reasonings from the text and determining their total complexity; 7) finding the share of logical reasoning in the text. The complexity of four text fragments was estimated, for which the share of logical reasonings and the average coefficient of information folding were calculated.