“Help with Learning Outcomes” (Hjálp með hæfniviðmiðin) is a custom AI chat in ChatGPT that supports teachers in formulating and revising learning outcomes, as well as linking them to assessments. Among other things, it can help avoid common mistakes when writing learning outcomes, such as using non-measurable verbs.
The answers or advice provided by “Help with Learning Outcomes” are suggestions that the teacher must evaluate and adapt to their own course. The teacher always carries the professional responsibility for what is ultimately published about the course.
“Help with Learning Outcomes” is designed for Icelandic, but can respond in other languages if requested.
Conversation starters
Users can choose to start a conversation from a conversation starter. Three starters are available:
- Create learning outcomes for a course (Búa til hæfniviðmið fyrir námskeið)
- What do you think about these learning outcomes? (Fá umsögn um hæfniviðmið)
- How do I link assessment to learning outcomes? (Að tengja námsmat við hæfniviðmið?)
After clicking on a starter, information about the next steps appears, along with a brief overview of the theoretical foundations on which it is based. The user is asked to provide the elements needed, for example, to create learning outcomes for a course.
The more detailed information the teacher provides in the chat — e.g., about the course, assignments, assessments, teaching emphases, type of competence, or other relevant factors — the more realistic and targeted the answers will be.
It is also possible to talk to the chat more generally about issues related to teaching and learning.
Bloom, SMART, and Biggs
Theoretical foundations:
• Revised Bloom’s Taxonomy (2001), i.e., the cognitive, affective, and psychomotor domains.
• Use of measurable action verbs according to Bloom.
• SMART quality criteria (Specific, Measurable, Achievable, Relevant, Time-bound).
• Constructive Alignment (Biggs).
In the chat, the user can ask for further explanations of these foundations. It is not necessary to know the underlying theory to use the chat.
Some examples of use
“Help with Learning Outcomes” can also be used for other teaching- and learning-related purposes, e.g., to speed up teachers’ work and as an idea generator. A few examples:
- Receive suggestions for assignments, teaching methods, and assessments tailored to specific parameters.
- Ask the chat to create a rubric for a particular assignment or learning outcome.
- Have the rubric generated for Canvas (as a CSV file that can be uploaded).
- Get ideas for assessment methods that align with a specific learning outcome.
- Ask for lists of action verbs, based on Bloom’s taxonomy, by domains and levels.
- Get explanations of specific teaching methods, etc.
Data protection
Sensitive personal data (such as national ID numbers and health information), unpublished research data, or confidential material should never be entered into the AI chat. Conversations may be stored temporarily in OpenAI’s service environment.
The chat only uses information that appears in the current conversation; if the user starts a new chat, it has no access to earlier conversations.
The author of the chat does not see conversations and receives no information about individual teachers’ use.
Limitations and further development
Since AI is still in its infancy in Icelandic, occasional language errors and awkward phrasing may appear in the answers. It’s also possible that the AI may sometimes invent information that isn’t correct.
“Help with Learning Outcomes” is still under development, so it would be very valuable for the author to receive feedback and brief comments on its use (please send to kriol@hi.is).