Dataset
GPT_raw_output.zip - contains all the original text files output by ChatGPT 4.o in response to essay-question prompts. 

There are 7 folders, labelled according to the question type in our taxonomy.

LQ = long quotation
Q = quotation < 40 words
NAS = not author specific
AS = author specific
O = options
NO = no options
MO = multiple options

Each folder includes 15 generated outputs.

Software
textstat.ipynb - calculate Fleisch-Kincaid readability scores for each question output
parse_sentence_lengths.ipynb - calculate the number of words, sentences, and paragraphs in each question output
parse_keywords - find designated keywords and identify the titles of literary texts named in outputs