Overview

There is a quiet question behind every module you build: how do you actually know a learner learned it, and how do you respond so they get better? This module answers both. On the Academy the answer is never "they watched the lesson", it is always an artifact a learner produces, a method that verifies it, and feedback that tells them what to do next.

We cover the whole loop: choosing a verification method, writing a rubric, building a quiz, and giving feedback that moves a learner forward. And you do not learn this alone. XRPL Commons offers 1:1 coaching to new content creators, and the review your own drafts get before they publish is feedback in action, the same craft, pointed at you.

The five verification methods

Every assignment carries one verification method, and choosing it is the core assessment decision:

self: the learner confirms their own work against a clear standard. Right for low-stakes practice and foundational checks.
peer: a classmate reviews the artifact against your rubric. Right for applied work where a second perspective adds value.
mentor: an instructor reviews it. Reserve it for high-stakes or hard-to-judge work; it is the most expensive verifier.
external: the platform checks the learner's own on-ledger activity automatically. Right when "done" is a fact on the testnet.
quiz: an auto-graded set of questions. Right for a fast check that the foundations stuck.

You set the method on the assignment's Details tab (see Hands-On in the Studio). Match it to what the task asks and to the stakes: a warm-up does not need your eyes; a capstone artifact does.

📸 Screenshot: the assignment verification-method picker.

Assess to steer, and assess to judge

The most useful distinction in assessment is why you are assessing at this moment. Formative assessment happens while learning is still in motion; its job is to inform, so the next step can adjust. Summative assessment happens at a boundary; its job is to judge whether the standard was reached.

Most courses are over-weighted toward summative and starved of formative. That is backwards. If the only time you learn that half the room misunderstood trustlines is a final check, you found out too late to teach. Build frequent, low-stakes formative checks into the flow, and reserve summative judgement for genuine boundaries. Quizzes and self-checks are your formative tools; a mentor-reviewed artifact or the journey capstone is where summative judgement belongs.

Rubrics a second reader could trust

A rubric turns a vague intention ("good work") into a judgement someone else could reproduce. It has two axes: the criteria (the few things that matter for this work, correctness, clarity, justification) and the quality levels (named bands like excellent, good, needs work, incomplete, each with a concrete description).

Three habits separate a rubric that works from one that just looks official: make each criterion observable (something you can point at on the page, not "shows good understanding"), describe each level rather than only labelling it, and share it before the work starts so it teaches learners what excellent looks like. Test it the way you test an exercise: give the same work to a colleague and see if your grades match. This rubric is exactly what a peer or mentor reviewer applies in the Reviews queue, which is where feedback happens next.

Give feedback that moves a learner forward

When a learner completes an assignment set to peer or mentor verification, their submission lands in the journey's Reviews queue. You open it and either approve it or return it with a note. The note is where the teaching happens.

Good feedback is specific (it points at the actual work, not a generic remark), actionable (the learner finishes reading and knows their next move), and tied to the criteria (it connects to your rubric, not your taste). Lead with the one change that would most improve the work rather than an exhaustive list a learner will ignore. When the issue is one person's rather than the whole cohort's, a direct message reaches them without an audience, which is gentler and usually lands better. A grade tells a learner how they did; feedback tells them what to do next, and only the second one moves learning forward.

Quizzes that grade themselves

The quiz method is the one piece of automated assessment the platform gives you. A quiz is a question bank: each question is single- or multiple-answer, with the correct option(s) marked; you set a pass percent and, optionally, a draw count so each learner sees a different subset. The correct answers stay on the server, so a quiz is safe to use as a real check, and a passing quiz completes the assignment on its own. Use quizzes as fast, low-stakes, formative checks, and let projects and artifacts carry the higher-order judgement. You build them on the Studio Quiz tab (see Hands-On in the Studio).

Close the loop: the Check and the feedback survey

Every module you write ends in a Check: a short section that names the artifact a learner produces and how it is verified. Writing it is how you make "done" observable instead of vague. When you can state a module's Check, outcome, artifact, method, you have designed its assessment.

The loop closes the other way too. The journey's feedback survey surfaces around the midpoint and reports a recommendation score, a rating of the mentors, whether the pace was too easy or too hard, and an open note. Read it as data, not a verdict: a low pace score and a confused note tell you exactly where to slow down or add a worked example next time. Treating your teaching as something to refine, rather than a finished artifact, is the same open-education habit your learners will copy from you.

Check

You are ready when you can design an assessment for a real outcome of your own. Take one learning outcome and write a one-page plan: state the outcome, choose a verification method and say why it fits, decide whether you are assessing formatively or summatively, write a rubric with at least three observable criteria across named quality levels, and give one example of the actionable feedback you would leave. Hand it, with a sample piece of work, to a peer and have them grade against your rubric. If their judgement matches yours, your standard is shared rather than private, which is the whole point.

Mentorship & Evaluation