Analyzing linguistic features of a given text like a thesis is a popular mode of investigation in which the researcher seeks to describe the content, structure, and the function of messages conveyed by the respective author. Generally speaking, there are various approaches to doing text analysis addressing micro or macro-levels of discourse (i.e., register or genre). In your case, we want to investigate the micro-level linguistic features of the targeted text (e.g., the thesis). For this purpose, we should first identify the unite of analysis (e.g., type of thematization, meta-discourse markers, features of text complexity, moves comprising different sections and subsections of the text). Finally, we should use a nominal/ categorical scale for analyzing the tokens belonging to each particular type. For providing you with a typical example , I refer you to an article titled "Sentence Initial Conjunctive Adverbials in Academic Articles Written by Native English Speakers and Iranian EFL Writers" available on researchgate.