Adaptive Learning Technology in the English Language Arts Classroom
How We Can Foster Writing Skills with Innovative Software
The ability to craft an articulate, engaging, and clear piece of writing can make or break a student’s opportunity to convey arguments, collaborate with peers, or make their voice heard on a large scale in communal spaces. However, writing instruction hasn’t always been associated with the same rigor or sustained practice as mathematics education, or even literacy skill-building. As the field of education technology has expanded, we’ve explored the implications of using adaptive learning software in the areas of math and literacy. But, what if we viewed writing instruction through that same lens of intensity, and implemented adaptive software in ELA classrooms?
The researchers at Stanford University — the minds behind Redbird Mathematics — have developed technology that will enable ELA teachers to drive instruction through advanced digital learning. This new software automates writing instruction, a feat that researchers have previously regarded as a daunting challenge because of the complexities of language. By building software driven by a stochastic motion engine developed at Stanford University, this adaptive ELA program provides steady, guided practice, detailed feedback, and personalized pacing of individualized instruction.
For students, having access to adaptive tools like this one could be a game-changer in writing instruction. As the student progresses through a lesson, adaptive software evaluates mastery of a concept, and presents a new task based on what that individual student is ready to learn next.
We’re constantly working to improve our adaptive learning software for ELA instruction. Our partnership with Stanford University allows us to conduct research that draws from a student composition database of ten million individual sentences and several hundred thousand paragraphs. Here’s a snippet of the initiatives the Stanford team is pursuing, all of which are aimed at improving the precision and breadth of error analysis in student writing.
Paraphrase-based detection of semantic errors
Currently, the software provides evaluation and explanation of student errors in grammar, meaning, and style. This research seeks to expand analysis of semantic errors (syntactically correct, wrong meaning), through adding “systematic treatment of paraphrase equivalence.” This effort draws on knowledge about semantics, logic, grammar engineering, and language processing.
Paragraph composition: Evaluating content and style
In its current state, the software analyzes sentence-specific errors, but there’s room for growth in analyzing errors that involve multiple mistakes in a paragraph. This work draws from linguistic research grounded in student composition data. It also examines effective user interface design to communicate error analysis to students.
Evaluating short essays written with open vocabulary
Lastly, researchers are working to develop lectures and exercises teaching the composition of short essays, rather than just vocabulary-bound sentences and paragraphs. This research involves the adaptation of methods for doing logical inference and knowledge representation.
Find a deep dive of the research and technology in this white paper by Dr. Dan Flickinger, Senior Research Associate at the Stanford University Center for the Study of Language and Information: