This research was conducted as a case study, which is one of the qualitative research methods, to determine how prospective Turkish teachers explain the concepts in their minds about writing skills. The study group consists of 60 3rd-grade prospective teachers studying at the Department of Turkish Language Education of Hatay Mustafa Kemal University in the 2018-2019 academic year. Research data were collected through a word association test. In order to reach the concepts that best describe the writing skill, first of all, ten concepts were determined by the researcher within the scope of the relevant course. Then, the test prepared with these concepts was presented to the expert opinion, and the word association test was designed with the remaining six concepts (writing, text, text types, narrative text, informative text, and argumentative text). In the data analysis, the answers of the prospective teachers to the word test were examined in detail and independently with an expert to ensure reliability. Frequency distribution was created with the frequency of repetition of the concepts related to key concepts. According to the frequency of the concepts, the breakpoint technique (Bahar, Johnstone & Sutcliffe, 1999) was used. In this study, breakpoint points between 20 concepts and above, 15-19 concepts, 10-14 concepts, and 5-9 concepts were applied. Concept networks were created separately according to the pre-test and post-test. According to the data obtained, the concept in the pre-test (f=1.636) increased to (f=1.732) in the last test. While the least increase was observed in the "writing" key concept (f=299-302), the highest increase was found in the "text" concept (f=263-292). The number of sentences containing academic information about key concepts increased from pre-test to post-test (f=116; f=179). The number of sentences without academic information (f=189-f=115) decreased. Moreover, the decrease in sentences containing misconceptions (f=29-f=13) supports this situation. When the sentences were examined in terms of including scientific information, it was found that academic expressions were used more accurately, and misconceptions were reduced.