It’s very possible that programs can be biased if you don’t train them correctly. AI algorithms help grade the ETS’s GRE and other standardized tests. Subscribe to Reset on Apple Podcasts, Stitcher, Spotify, or wherever you listen to podcasts.Īrielle Duhaime-Ross spoke with Aoife Cahill, a managing senior research scientist at Educational Testing Service. In addition to Feathers, you’ll hear from Utah parent David Hart Aoife Cahill, a managing senior research scientist at Educational Testing Service and Vox reporter Sigal Samuel. Below, we’ve also shared a lightly edited transcript of the episode. Listen to the entire discussion on this episode of Reset. And once those algorithms are built, explains Reset host Arielle Duhaime-Ross, they can reproduce those biases at a huge scale.Īnd the worst part? You can’t cross-examine an algorithm and get to the bottom of why it made a specific decision. Depending on the program, those predictions can be consistently wrong in the same way. Then, the company trains an algorithm to make predictions as to how a human grader might score an essay based on that data. There are a bunch of different algorithms, made by a bunch of different companies.īut they’re all made in basically the same way: First, an automated scoring company looks at how human graders behave. The reason it’s so hard to figure out who’s affected by AI grading is because there’s not just one program that’s being used. But here’s what we do know: These programs are being used to grade students of all ages and levels, from high school students to students applying to grad school, from middle school students even down to those in elementary school. And the other problem is that some of the algorithms have been proven by the testing vendors themselves to be biased against people from certain language backgrounds.”įeathers wasn’t able to pin down exactly how many students are affected by this. It looks good from afar but it doesn’t actually mean anything. One is that they can be fooled by any kind of nonsense gibberish sophisticated words. “The algorithms are prone to a couple of flaws. Todd Feathers, who wrote about AI essay grading for Motherboard, called up every state in the country and found that at least 21 states use some form of automated scoring. So can artificial intelligence really teach us to write better? Algorithms are grading student essays across the country.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |