This thesis investigates linguistic data and how it is processed using the example of children's speech. During the acquisition of the first language, the child goes through certain stages of language development, about which today much is known precisely because of the methodology of collecting and processing children's speech. Due to the development of electronic devices, it is easier to access data and create corpora that help explain certain language phases of a child's speech. One of the ways of transcribing and processing data is the CLAN program package, which was developed for the purposes of the CHILDES database. The aim of this paper is to apply this model and its norms in the transcription of audio recordings of the child's speech production in Italian as the target language, and to explain the advantages and disadvantages encountered during the process. Analyzing the audio recording in the CHAT program provides information about the codes used and their importance for the research of children's speech. The analysis showed that the CHAT program is a useful tool for data processing. The program offers many functions thanks to which it is easier to gain new knowledge about the development of children's speech, speech errors and code-switching in the case of bilingual children.