Apparently, Data Driven Learning Related To Language Learning Doesn’t Mean What You Think It Does

akshayapatra / Pixabay

I’ve never been a big fan of being “data-driven” in education (see The Best Resources Showing Why We Need To Be “Data-Informed” & Not “Data-Driven”).

Imagine my surprise, however, when I learned that the term “data driven learning” apparently means something completely different in language learning circles.

I saw a study titled The impact of data-driven learning form-focused tasks on learners’ task engagement: an intervention study.

It sounded like data-driven learning was awfully similar to inductive learning (see The Best Resources About Inductive Learning & Teaching).

So, I asked ChatGPT:

Is data-driven learning (DDL) similar to inductive learning?

Here’s its answer:

Yes. Data-Driven Learning (DDL) is closely related to inductive learning, but they aren’t synonymous. I’d describe DDL as one particular way of doing inductive learning, especially in language teaching.

In inductive learning, students examine examples and try to infer a pattern or rule rather than having the teacher give them the rule first:

Examples → notice patterns → formulate rule → test/use rule

In Data-Driven Learning, the examples—the “data”—typically come from a language corpus, meaning a large collection of authentic language. Students examine multiple examples of a word, phrase, grammatical construction, collocation, etc., and look for patterns.

For example, rather than telling English learners:

“Interested” is normally followed by “in.”

you might give them corpus examples:

She’s interested in science.
They’re interested in learning English.
I’m interested in the position.
He became interested in photography.

Then ask:

“What word usually comes after interested? What pattern do you notice?”

The students discover interested + in, which makes this both DDL and inductive learning.

The important distinction

You can have inductive learning without DDL. A teacher might create five sentences illustrating the past tense and ask students to determine the rule. That’s inductive teaching, but it isn’t normally considered DDL because students aren’t investigating corpus/language data.

And DDL doesn’t have to be completely inductive. A teacher could tell students a rule first and then have them use corpus data to investigate whether, when, and how the rule operates. That’s DDL being used more deductively.

So I’d represent the relationship roughly as:

Inductive learning = the broader instructional approach
DDL = learners acting somewhat like “language detectives,” using authentic language data to discover or investigate patterns.

For ELL teachers, I think “inductive learning using authentic language data” is probably the clearest one-sentence explanation of DDL.