Educational analytics applied to lexical errors in foreign language writing

educational analytics is a term mostly used in the context of university management, although the use of big data in education also affects the teaching of individual academic subjects. This article analyzes data from the Petrozavodsk Annotated Corpus of Student Texts (ПАКТ), focusing on lexical errors made by students writing in German as their second foreign language after English. The analysis is conducted in several aspects: quantitative, substantive, potential causes of lexical errors, and the degree of their severity. It concludes that the number of lexical errors in student texts in German significantly exceeds the amount of all other types of errors. During the study, 11 groups were identified that more precisely describe the nature of lexical errors. Relying on statistical data from the corpus, the authors confirm the thesis that lexical errors, more than others, affect understanding, i. e., they impede communication to varying degrees or make it completely impossible. Furthermore, the study concludes that interference from the native language has a greater impact on the occurrence of lexical errors than misprints or interference from the first foreign language.