Analysis of science textbooks and supplementary resources using word frequency lists

This study aims to determine the frequency distribution of the words and to discuss the similarities and differences in science concepts and science teaching in the 8th-grade science textbooks accepted as textbooks by the Ministry of National Education and supplementary resources prepared by different publishers. The methods in natural language processing and text mining were used for the study. The texts in the textbooks and reference books were converted into plain text, and the frequencies of the words in the relevant textbooks were reached with the help of NKod software prepared in JAVA programming language using the Zemberek NLP Application Programming Interface. R Programming Language (R Core Team, 2021) and RStudio (RStudio Team, 2020) applications were used for data analysis. The findings obtained as a result of the analysis are as follows. - It was found that the first 10 and 100 terms used in the course and supplementary resources reflect the unit's content. In addition, the most emphasized terms in the course and supplementary resources are generally the same. - The highest Spearman's rank correlation values between the units of the books belong to both "Pressure" and "Simple Machines" and "Seasons and Climate" and "Energy Transformations and Environmental Science" units, and the lowest correlation values belong to the correlation value of "DNA and Genetic Code" unit with "Matter and Industry", "Simple Machines" and "Electric Charges and Electric Energy" units. - The highest Spearman's rank correlation values between the same units of the books examined in the study belong to the auxiliary sources, while the lowest correlation values belong to the correlation values between the textbooks and the auxiliary sources. - The statistical difference between the frequency distributions of the terms at the unit levels was examined with the chi-square test, and it was seen that the statistical differentiation between the official textbooks was low. In contrast, the statistical differentiation between the textbook and the source books was high. Moreover, it was found that the most statistically differentiated terms for all units except the "Electric Energy and Electric Charges" unit were the terms belonging to the unit's content. According to the statistically differentiated terms, the contents emphasized by the textbooks were interpreted, and it was found that, in general, the sourcebooks emphasized questions and sometimes non-acquisition terms belonging to the unit. In the activities, students tried to reach the results through experiments and observations. In textbooks, on the other hand, it was found that there were fewer questions, terms related to the activity were emphasized rather than terms related to the unit, and ready-made experiments were given to students in the activities.