I was recently asked a very good question about how to change one’s assessments to align more with CI. By that, we’re talking about comprehension-based language teaching (CLT) that prioritizes comprehensible input (CI) in the Latin classroom. First, it helps to think in terms of what standards were being assessed beforehand, even if they weren’t explicitly called “standards.” These old standards were mostly discrete skills you’d expect to find in tests accompanying popular textbooks, like vocabulary recall, derivative knowledge, grammar identification, and cultural trivia. New standards based on CI—whatever they are—will have meaning at the core. My suggestion is to focus on assessing comprehension of Latin, because that’s more than enough to ask for. One benefit of this standard is that is that it has those old discrete skills embedded within something larger and more meaningful that you can assess (i.e., comprehension). Let’s look at how each one of the old standards is contained within assessing comprehension…

Vocabulary Recall
This is the easiest one. Vocab quizzes are a waste of time. They might show basic recall, sure. Or, they might just show short term ability to cram for a quiz! You can do better. Instead of assessing isolated vocab words, assess understanding of Latin in context and it will inherently contain the vocab words you would’ve tested anyway.

CI Assessment Example:
If your old vocab quiz had est, sunt, and et, rather than having a fill-in-the-bank or matching section quiz, write a couple sentences that include the vocab. For example, write Boston in Massachusetts est. Burlington et Brattleboro in Vermont sunt. Then, ask—always in English—“where is Boston?” Or, ask “what cities are in Vermont?” Maybe even ask “which word means ‘are,’ sunt or et?” These are assessments, and the responses to these questions show you understanding of Latin. Even the direct “what does this word mean?” now has actual context in the form of complete sentences. These sentences used to assess can be shared digitally, printed out, or projected in front of the whole class. I preferred the latter, making assessment PART of the learning process during class. Paragraphs are even better: project a paragraph, have students read for a minute, THEN ask a comprehension question. This gets students reading (i.e., more input) since they aren’t just skimming for clues in the question prompt, known as “backpedaling.” You even don’t need to write your own Latin, either. Maybe start by taking existing text that contains vocab you would’ve quizzed in an isolated way, like textbook passages that include the vocab anyway.

Derivative Knowledge
This one’s trickier, but any time a student uses derivatives and cognates as clues to understanding, and you assess that understanding, you’re also assessing their knowledge of derivatives. It just might be unobservable. You could certainly make that more visible by asking students to tell you how the words themselves aided comprehension (e.g., “were there any clues in the words in this sentence/paragraph that helped you understand?”). Of course, how much does that matter if the goal is reading and understanding Latin?! Sure, breaking down words into smaller parts is a nifty strategy for comprehension (i.e., one step along the way to arrive at something greater, understanding Latin).

CI Assessment Example:
If your old assessments had derivative sections, and you really want to assess how that knowledge helps students read, start with a paragraph of Latin:

ego nōn vīdī omnia sīdera. nōn omnia sīdera in cōnstellātiōne Ōrīōne illūminābant. cūriōsē, ego haud potuī vidēre cingulum Ōrīōnis. nūllum sīdus in cingulō vidēbātur—impossibile! Ōrīōn cingulum habuerat. Ōrīōn habuerat cingulum aetātēs multās. sed iam, Ōrīōn cingulum nōn habuit. post aetātēs multās, cingulum Ōrīōnis ēvānuit!?  mystērium, et quoque ōminōsum erat…

Then, have one of the prompts be “underline all the Latin words that became English words” or something, and then “choose five of those Latin words you underlined, and write an English word that came from each one.” Finally, ask a comprehension question—because that’s the real standard here (i.e., understanding Latin), and this will be more aligned with CI.

Grammar Identification
Rather than testing grammar identification, any test of comprehension can show whether the grammar contained in Latin words is being understood. For example, specific prompts about when something occurred can tell you if a student is understanding tenses. And if they don’t, that might be fine. Rather than testing precise use of language, consider ballpark comprehension in the beginning years. Even college students mix tenses in their native language, right?! Surely, grammar knowledge isn’t needed to read and understand Latin, so maybe this has no place being assessed anymore, even if it’s embedded within certain comprehension questions.

CI Assessment Example:
If your old assessments had grammar questions, convert them into comprehension questions highlighting the grammar function. For example, start by projecting a paragraph:

Architectī custōdēs sīderum erant. Architectī custōdiēbant sīdera post Cōnflīctiōnem Necessāriam. observātiōnēs Architectōrum Aetāte Septimā (VII) extraōrdināriae sunt. nam, ēvidentia Architectōrum ēvānuit Aetāte Septimā (VII). post Aetātem Septimam (VII), nūlla ēvidentia Architectōrum inventa est.

Asking “how are the observations described?” would be an adjective question. Asking “whose evidence vanished?” would be a genitive case question. Asking “what was not found?” would be an adjective, genitive, and perfect tense question. And if you’re thinking “bUt ThOsE aRe JuSt vOcAb QuEsTiOns?!” then congratulations! Words contain grammar, and are inseparable in actual use.

Cultural Trivia
This one is so easy it hurts. Rather than testing students’ ability to memorize all the facts about the Romans, etc., ask comprehension questions about Latin that have cultural topics embedded within. That way you’re still assessing the most important standard: understanding Latin. The content of Latin is secondary in this case, and one could argue that being able to read and understand a variety of Latin Latin texts is the real point, anyway. Otherwise, we’ve got plenty of history books in English, and plenty of history courses on Roman culture.

CI Assessment Example:
If your old assessments had tidbits about the Romans, just use Latin paragraphs about Romans and ask comprehension questions. Literally anything. Here’s one on provincial Egypt:

Aegyptus erat prōvincia Rōmae (id est, pars Rōmae). Aegyptiī rēgnābant (exemplī grātiā, Cleopātra). sed, Rōmānī rēgnābant Aegyptiōs. ergō, Rōmānī in Aegyptō habitābant. in pictūrā sunt Sphinx et pȳramidēs. Sphinx et pȳramidēs in Aegyptō erant. in pictūrā sunt quīnque obeliscī. obeliscī variī sunt. sunt obeliskī immēnsī, et sunt obeliskī minimī. quoque sunt statuae animālium. sunt canēs minimī, et Sphnix. Sphinx fēlēs vidētur esse. in pictūrā, Sphinx cōnstruitur. sed, ubi sunt hominēs?! est vir sōlus in pictūrā?! vir vidētur esse Aegyptius. habetne vir amīcōs? triste!

Ask “how is a province of Rome described?” or “who ruled whom?” or “what objects are found in the picture being described?” Any comprehension question will assess both understanding of Latin and cultural knowledge.

WAYS To Assess Comprehension
So, you’ve now got a new comprehension standard that aligns with CI. There are three ways to go about assessing that:

• Product – WHAT students understand
• Process – WHAT they DO to understand
• Progress – HOW MUCH they’ve improved their understanding

All of the CI Assessment Examples above are products. While product is the classic, straightforward choice, I advocate for self-assessed process and progress, explained in this post. In the meantime, start by assessing products since it’s the most familiar. These products can be collected physically or virtually, either to review after being turned in, or in real-time, such as eliciting student responses. I advocate for the latter, asking a LOT of questions that get you data as to whether students are understanding Latin. You don’t even need a quiz. In fact, I say skip it! You’ve got highlight your confusion, or the monitor assessment options right there.

I can predict what’s coming next, though, and that’s what to do about grading.

Remember that grading is something different from assessing when it comes to learning. Grading is a summative function of reporting achievement. However, a) not grading/scoring, b) providing feedback, and c) involving the student in that feedback are all formative functions that improve learning. In other words, grading has no place during learning. CI isn’t much good if students are still getting points taken off for silly mistakes and errors we actually expect as part of that learning process. If you’re going to be updating your assessments, you’ll likely have to update your grading practices. This is good.

The quick fix is to remove all points on all assessments whatsoever, instead using a holistic rubric, or simple criteria that you can provide feedback on. Keep a record of things students are DOING and submitting in your gradebook, but leave them all as collected only (i.e., no numbers or scores). Have students self-grade at your progress report time, again using holistic rubrics or simple criteria, and otherwise spend the majority of time providing feedback, which can be done asynchronously over Google Classroom, OneDrive, or email. For feedback, I recommend Mark Barnes’ SE2R model:

Summarize (e.g., “You wrote a one-page summary of the topic.”)
Explain (e.g., “You defined the similarities between X and Y. I didn’t see any statements to support your claim, though. Do you see them?”)
Redirect (e.g., “Add supporting statements and resubmit.”)
Resubmit (e.g., “When you finish, send me an email.”)

Understanding Latin is plenty of work, so it’s really the only standard you need. Maybe you want to grade WHAT students understand (i.e., products). Maybe you want to grade WHAT they DO to understand (i.e., process), or maybe you want to grade HOW MUCH they’ve improved their understanding (i.e., progress). If you choose just one of those, the grade is 100% of that. This is simple gradebook stuff, and simple to collect data on. You’ll never be left with not enough learning evidence (vs. if you had 5 categories, needing 3-5 pieces of evidence for each one to have enough data for trends). If you choose two ways to grade, they can be weighted 50% equally, or one weighted more depending on what you value. For example, if the idea of students self-assessing and self-grading their process of learning Latin is scary to you and feels too progressive like Bernie Sanders, instead, you can weight product 75% and process 25%, arriving to the right of center in more of a Biden approach. Maybe progress has no place in your schema. If so, don’t consider grading it all. When grading process, also be VERY careful that you’re not just listing bias-ridden criteria. When in doubt, use mine (from other posts above). They’re essential processes that humans NEED to go through in order to receive input. Whatever you decide for grading, use your professional judgment on this one. It’s already going to be better than the old ways.