AI and evaluation

This year, there are lots of interesting discussions on the use of artificial intelligence in evaluation. I have started trying out ChatGPT to help me find out about mainstream information/opinion on the subjects I have worked on, develop workshop plans, and draft an email message. The results are interesting: at first sight, ChatGPT responses look quite good, but closer scrutiny makes them seem quite mediocre. That is normal, because ChatGPT relies on what is out there – all of it, or an average view of all of it, i.e., not the most advanced or the most exciting thoughts. Still, one can build on a medium useful response to develop something more suitable for the task at hand. An important first step is to question what the machine has produced and to reflect on what is missing in the machine generated answer. It could be the entire construct the answer is based on…

Regardless of more serious applications, ChatGPT has tricks to cheer up our work routines. This is what it has produced when I asked for a summary of principles of evaluation in the style of Emily Dickinson:

To assess, to know, to understand,
The ways of evaluation, so grand.
First, set clear goals for what you seek,
And measure progress, week by week.

Gather data, both near and far,
To see what’s working and what’s not par.
Look for patterns, trends, and clues,
And use this knowledge to inform your views.

Consider all perspectives, wide and narrow,
To understand the impact, deep and thorough.
Be open to feedback, both good and bad,
And use it to improve, like iron clad.

Celebrate successes, big and small,
And share them widely, one and all.
Acknowledge challenges, too, with grace,
And work together, to find a new pace.

In all your evaluations, always strive
To keep an open mind, and be alive.
For learning and growth, and progress to come,
Are the gifts of evaluation, to some.

Add a Comment

Your email address will not be published. Required fields are marked *