Table 3 Encoding framework for interaction with ChatGPT.
Category | Code | Description | Sample |
|---|---|---|---|
knowledge inquiry (KI) | Irreverent knowledge inquiry (IKI) | Students raise questions unrelated to the topic. | What historical initiative should college students possess? |
Superficial-level knowledge inquiry (SKI) | Students raise questions related to the topic, but without personal processing or perspectives | Python code for creating radar maps. | |
Medium-level knowledge inquiry (MKI) | Students raise questions related to the topic, which have been processed by themselves | Create radar images using some libraries in Python and provide detailed code. | |
Deep-level knowledge inquiry (DKI) | Students raise questions related to the topic, clearly articulate their viewpoints, arguments, or opinions, and provide detailed problem-solving steps | Using the Numpy and Matplotlib libraries in Python to create a Holland career analysis radar map, it is necessary to include 6 labels and provide accurate and detailed code that can run. | |
Feedback inquiry (FI) | Superficial-level feedback inquiry (SFI) | Student simply (dis)agreement, or simple evaluation for the feedback from ChatGPT. | Please continue writing. |
Medium-level feedback inquiry (MFI) | Students evaluate ChatGPT’s feedback and ask for more simple questions. | What does linewidth=2 mean? | |
Deep-level feedback inquiry (DFI) | Students evaluate the feedback on ChatGPT and provide specific examples for further information. | After entering the answer from you, there were errors such as Value Error: x and y must have the same first dimension, but have shapes (6,) and (5,). Please provide a modification plan and possible code. | |
Feedback from ChatGPT (FC) | General feedback from ChatGPT (GFC) | ChatGPT returned general feedback without specific problem-solving steps. | As a virtual assistant, I am able to provide you with help and support in Python programming. I don’t have an actual identity, but I can help you like a programmer who is familiar with Python. |
Specific feedback from ChatGPT (SFC) | ChatGPT returned examples of text and code, and it has a clear structure or block. | Ok, taking Weibo as an example, the process of crawling resources with the theme of “postgraduate entrance examination”, “English”, “mathematics” and “tutorial” is roughly as follows: 1… 2… | |
Correct feedback from ChatGPT (CFC) | ChatGPT returned correct feedbacks of the questions raised by students. | Ok, first you need to install Numpy and Matplotlib, which you can do with the following command: 1… 2… | |
|  | Incorrect feedback from ChatGPT (IFC) | ChatGPT returned incorrect feedback, such as incomplete answer or server error. | Ok, taking microwave as an example, the process of crawling resources with the theme of “postgraduate entrance examination”, “English”, “mathematics” and “tutorial” is roughly as follows: 1. Identify sites to crawl: Yes |
