AI chatbot user survey

In light of the growing presence of AI in education, Leiden University is exploring how this technlology interacts with educational activities. The purpose of this survey is to provide an overview of how teachers and students utilise Large Language Model (LLM) chatbots, such as ChatGPT, Gemini, Claude, or Microsoft CoPilot, among others, to support a safe and effective implementation of AI in education.

For the sake of the questions in this survey, consider activities such as preparing classes, writing assignments, designing or adjusting assessment rubrics, providing feedback to students, or a combination of the above.

This survey is anonymous and takes approximately 7 minutes to complete. Its results will be used to advise the university's administrative board on the preferences and needs of the Leiden University community.

If you are interested in more information about the survey and its results, please contact the Leiden Learning and Innovation Centre.

What course level do you primarily give classes in? *

Bachelor

Master

In what faculty are you working as a teacher? *

About your use of LLM chatbots

Do you use a LLM chatbot for your tasks as a teacher? (E.g., ChatGPT, Gemini, Claude, Microsoft CoPilot, etc.) *

Yes

If you do not use an LLM Chatbot for teaching activities, please explain shortly why

Please select in what activities you use an LLM chatbot (you may select multiple) *

Preparing classes

Researching information for course content

Designing assignments

Providing feedback to students

I don't use a LLM Chatbot for any activity related to my role as teacher

Other:

What language to you use when interacting with an LLM chatbot for education activities? *

Did the level of transparency of the LLM chatbot influence your choice? *

Yes

Did not consider

Are there other aspects you take in consideration when choosing an LLM chatbot for your tasks as a teacher?

About your preferences

This section covers your hypothetical preferences when choosing a LLM chatbot.

Please rank what you consider most relevant, with 1 being most relevant and 6 being least relevant, when choosing an LLM chatbot for your studies. If you regularly only use one LLM chatbot, then rank them based on the attributes that would make you consider another.

You may only assign each rank to one attribute.

If you do not use an LLM chatbot for your studies, then rank each based on the importance you assign to each.

The chatbot accepts different types of input methods (E.g., text, voice, image) *

The chatbot provides different types of output methods (E.g., text, voice, image) *

The chatbot is fast when delivering output *

The chatbot is precise when delivering outputs (I.e., the output requires less editing or fewer additional prompts to arrive at a satisfactory output) *

The chatbot is accessible and/or easy to use *

The chatbot explains and/or justifies its output in a detailed manner *

How relevant is the way the chatbot is trained* for you?

About training: LLMs are a type of AI that are trained on massive amounts of data to produce human-like responses to natural language queries. This training data can include raw, text, audio, or visual data from web pages, forums, blogs, Wikipedia entries, books, articles, technical documents, or other internet-based sources. *

It is relevant

It is not relevant

I have not considered it

How relevant is the hosting* of the chatbot for you?

About hosting: Depending on the LLM chatbots, they are hosted at location with potentially differing data privacy regulation. *

It is relevant

It is not relevant

I have not considered it

* = Input is required