SI in a Real Classroom: What I Tried, What Worked, and What I'm Still Figuring Out

If you've made it to Part 3, you've done the reading. You know why students can understand input without processing it, and you know what referential and affective activities look like in practice. Now I want to get into what this actually looks like when you try to use it in a real class, with real students, on a real Tuesday.

This post is less a research summary and more an honest debrief. Here's what I've tried, what I think worked, what I'd do differently, and one idea for how SI principles can help even if your department is tied to common assessments.

The Gaslight: "If You Know the Research, Implementation Is the Easy Part"

It is not.

Knowing that learners process utterance-initial (beginning of a sentence) forms more easily than utterance-medial (middle of the sentence) ones is genuinely useful. Knowing it does not automatically make you good at designing activities that account for it. There's a gap between understanding a principle and building something that actually applies it, and I have definitely made activities that I thought were doing one thing and were doing something else entirely.

The Gatekeep: What I Actually Tried

My Spanish 4 students have a complicated relationship with ambiguity. They grew up with traditional assessments. They want something to study, something with the right answer, something they can feel in control of. For a while, I gave them content-based grades with no traditional tests, which I thought was reasonable and fair. And it was. But for many of them, the lack of explicit stakes meant a lack of attention.

So I started building unit content tests. Multiple choice, content-focused, something students could actually prepare for. And as I was writing the first one, I was also reading Farley's Structured Input, and I started noticing that the principles I was learning about for activity design applied just as much to writing good assessment items.

That's when things got interesting.

The tool that came out of that process is what I think of as a grammar-content matrix. For each multiple-choice item, I write four possible answers.

The correct answer required both correct grammar AND correct content. One distractor has correct grammar but wrong content, so a student who processed the form but didn't understand the reading lands there. Another has incorrect grammar but correct content, so a student who understood the reading but didn't process the form lands there. The fourth option is your catch-all.

From a data perspective, this is actually really useful. If a whole class is landing on correct grammar-wrong content, that's a content gap. If they're landing on wrong-grammar-correct-content, that's a processing gap. Those two problems call for different responses.

I also put what I want students to process at the beginning of each answer option, not buried in the middle. If I'm targeting singular/plural verb agreement, the verb comes first. Utterance-medial forms are the hardest to process, so hiding the thing I want students to notice in the middle of a long answer choice works against the whole point.

In the convivencia unit, this logic is evident in the workbook activities. The singular/plural activities ask students to add -n to a verb if the subject is plural and leave it blank if singular. To get it right, students have to read the full conditional sentence, identify the subject, and decide whether the verb agrees with a single entity or multiple entities. It's not "write the correct form." It's "understand this sentence well enough to know which form belongs here." That's the difference.

What I'd Change

A few honest things.

I leaned on pop-up grammar more than I needed to in the first year. Pop-ups have their place, but if I'm doing one because my activities aren't giving students enough reason to process the form, the pop-up is a workaround, not a solution. The activity design should be doing more of that work.

I also put some SI-style items on high-stakes tests that would have been much better as low-stakes in-class activities. The grammar-content matrix works well as a diagnostic and as a learning tool. It's less clear it belongs on a summative assessment. What I'm moving toward is keeping tests focused on content comprehension and using the grammar-content activities more frequently and at lower stakes throughout the unit, closer to when students are actually encountering the forms in input.

The convivencia workbook is the most complete version of this I've built so far. Every activity has a communicative Paso 2. The forms appear across multiple activity types, different verbs, and different contexts. Students aren't just answering comprehension questions or just doing form-focused tasks. They're doing both, in the same unit, with the same content, and then actually talking about it.

How This Could Help If You're Tied to a Textbook

If your department requires common assessments that include grammar, SI principles give you a way to make those assessments better without blowing up the whole system.

Instead of fill-in-the-blank conjugation questions, try forced selection or a word bank that already has the form students need. Take out the conjugation step and let them focus on communicating. The correct answer requires correct form AND correct comprehension. Distractors are meaningful wrong turns, not random wrong answers. Students who haven't processed the form can't guess their way to the right answer by eliminating obviously wrong options.

It's not a perfect solution, but it's a principled compromise. And it gives you useful data about whether students are missing content or missing form, which tells you something different about what to do next.

Wrapping Up the Series

Here's where I've landed after a few years of thinking about this.

Structured Input isn't a curriculum. It's a lens. It’s an intervention. Once you understand why students skip certain forms, you start seeing your activity design differently. You start asking whether a time marker is doing work that you meant for the verb to do. You start putting the form you care about at the front of the sentence. You start building follow-ups that require students to actually use the language rather than just complete the task.

There's also more to say about what happens after the input work is done, specifically, how to give feedback on form in a way that builds on what students have already acquired, rather than trying to replace acquisition with correction. That's a separate post, but it connects directly to everything in this series. Stay tuned.

The convivencia con animales unit linked below is my best attempt so far at putting all of this into practice in a content-rich, coherent unit. It's real, it's been used with actual students, and every activity in it is designed with these principles in mind.

If you try any of this, I'd love to hear what happens. Drop it in the comments, find me on Threads, or send a message through the podcast. That's what the scholar-practitioner community is for :)