In this section, I provide a transparent and reflective account of my use of artificial intelligence for my bachelor's thesis. The objective is to clearly communicate how AI tools were used and to what extent. 

 

Artificial intelligence - ChatGPT - was used for language refinement. As a non-native English speaker, I chose to write my thesis in English, as all used sources were in english. This helped to avoid unnecessary language switching and potential distortions in meaning. AI tools supported this process to improve the fluency and readability of my writing. All sentences were written by me first. ChatGPT was then used to suggest alternative formuliations with a more fluent writing style. No AI-generated text was ever copied and pasted directly into the thesis. Instead, the suggestions served solely as inspiration, which I manually reformulated in my own words. The purpose of this process was to improve clarity, correctness and readability while fully preserving my own words, structure and argumentation. 

 

I also used AI to support the process of shortening and condencing text. In several instances, passages I had written exceeded the required length or contained redundancy that needed to be reduced in order to align with formal requirements and word limits. In this context, ChatGPT was used, again, in an advisory capacity to suggest how passages could be reformulated more concisely without loosing substantive content. Again, no AI-generated text was copied directly into the thesis. The suggestions were only used as guidance and all final formulations were rewritten by me to ensure efficiency, coherence and consistency with my own writing style. 

 

The final use of artificial intelligence concerned the development of the Python script used for the semantic network analysis. Given the complexity of analysising legitimisation strategies in political discourse, I percieved semantic network analysis as the most suitable method to identify concepts and their interconnections. Python was, in this context, the most efficient and reliable tool for this task. As I had limited experience in programming, ChatGPT was used as an advisory tool during the development of the script. Its role was to provide guidance on how to structure the code, identify errors, explain why certain parts were problematic and suggest possible corrections. The development process of the Python-script took approximately ten days and involved repeated testing, debugging and manual correction. All coding was written and corrected by hand; with ChatGPT only providing assistance. 

Once the script was completed, it was additioally reviewed by a personal acquaitance with experience in coding, who confirmed that it fullfilled the analytical aim of the study. This ensured methodological reliability while allowing me, despite limited coding experience, to employ appropriate tools to pursue the research objectives of my thesis.