Abstract:

Data Driven Learning is a didactic approach which treats language as data, and which sees students as researchers doing guided discovery activities. Students can be researchers as they discover the new language using natural linguistic data. The branch of the scientific field of linguistics that focuses on corporal studies and applications, i.e. Corpus Linguistics, has brought about a new wave of corpus-based educational approaches in foreign language teaching. However, the exploration of corpora can effectively support reflection also on specialized languages, such as that of Mathematics. Using these didactic approaches, which also include the use of technologies, students develop digital skills, as well as linguistic and mathematical skills. This paper presents a corpus-based activity which involved 36 Italian students of grade 11 and their teachers. This 1-hour activity was proposed in the last meeting of an 8-hour experimentation characterized by Data Driven Learning activities on the language of Mathematics, in particular in the "Relations and functions" area. Students used the tool concordance AntConc (https://www.laurenceanthony.net/software/antconc/). This software makes it nice and easy for students to extract collocates and frequency lists, which makes it useful if one wants to explore the lexicon of a dataset and define the properties and relationships of a given word. In previous meetings the students had learned how to consult a corpus to carry out activities with automatic formative assessment in Italian. In this final activity, students were divided into small groups and they were asked to complete some activities in English, by consulting a corpus specially created by trainers. The first request concerned the creation of a word list (using the stop list provided), that is the search for the five headwords (singular form of the noun) most frequent in the corpus. If they did not know their meaning, they had to consult the corpus to find out. Then some linguistic activities were proposed, which the students had already carried out in the previous meetings in Italian (e.g. " which verbs have the name "function" as subject?"). Finally, the last request was to translate typical expressions of the mathematical language (e.g. "set of inputs") from the Italian language to the English language.

The research question are as follows. Did the students answer the various requests correctly? What kind of impact can this type of activity have on students? To answer the research questions, all the students' answers and their responses to the final questionnaire, completed at the end of the experiment, were analyzed. Students worked collaboratively in groups, even in hybrid teaching cases (with some students in class and others at home connected online) due to the pandemic situation. In some cases, the students did not consult the corpus to answer the questions but used online translators and made translation errors. This aspect was important to make students reflect on the importance of knowing the specialized language of mathematics and of sharing the language to know the meanings of objects and mathematical expressions. The students also reflected on the characteristics of the Italian and English languages and on the importance of using translation tools in a critical way. The students enjoyed the group activity and the data driven learning activities very much.

Keywords:

Data Driven Learning, Language for Specific Purpose, Secondary school, STEM Education.