AI for Educational Assessment
McKinsey and company, the famed consultancy, just issued a report on the economic potential of generative artificial intelligence. The authors estimated generative AI could add between $2.6 trillion and $4.4 trillion annually to the global economy. They estimate this impact would double if they included the impact of embedding generative AI into software currently in use today. For comparison, they mention the entire gross domestic product of the United Kingdom in 2021 was $3.1 trillion. Wow!
How great an impact might this technology have on education and educational assessment? John Spencer, an innovative teaching expert, interviewed Mark Schneider, an educational futurist, who described one application of AI to speech therapy.
Right now, speech therapists have to spend hours giving tests to figure out what kind of help a person needs. This testing can take a long time and lack challenge for a skilled therapist. But what if we had smart computer tools, combining AI and voice recognition, that could do these tests for us? These AI tools could give the speech therapists critical information they need to understand and tailor a diagnosis for their speech client.
Does that sound like some far future technology? Think about how the SoundHound and Shazam apps on your smartphone identify musical artists and song titles from only brief passages of music. These technologies are very similar.
Aided by these not-too-distant technologies, machines might take over the tedious and time-consuming portions speech assessment. This parallels the history of factory work during the Industrial Revolution. Then speech professionals could spend more time working with people more interactively and personally. They could then focus on the creative and most highly skilled aspects of their work. As the McKinsey report points out, this would also allow therapists to see more patients and provide higher quality of care.
Now, think about a similar technology to help teach reading skills. Early elementary teachers now have to spend hours testing each student one by one. Why not automate the mechanical portions of reading assessment? When teachers, like all humans, doing straightforward and repetitive tasks, they sometimes make mistakes. But AI doesn’t get tired. What if we used AI with voice recognition to test how well students can read? The students would take the test on a computer or a smart gadget, get a score right away, and see how it compares to their reading goals. This would give teachers more time to have one-on-one conferences with students and to work with small groups. Teacher could also adjust lesson plans and individual goals more easily with frequent, accurate, machine-aided feedback on student skill levels.
AI is not going to replace teachers in testing and assessment. Instead, we could use AI to do the parts of testing where people might make mistakes. Also, an AI tool could more easily keep pace with new research and new methods.
The benefits of AI-assisted teaching and assessment outweigh the challenges. What are we waiting for?
*****
If you would like to read more about how some creative high school students used sound recognition technology to help a marine biologist protect endangered leatherback sea turtles, check out my book Teen Innovators: Nine Young People Engineering a Better World with Creative Inventions.
The McKinsey report on the economic impact of generative AI.
John Spenser’s May 22, 2023 blog, well worth reading.