Grading and Score

Berdee lets you decide what counts as a graded response, when grading happens, and where that result should matter.

This flexibility is one of the product’s strengths, but it works best when you think in layers.

The grading layers

There are three useful grading ideas to keep straight:

1. Widget grading
2. Scene score
3. Lesson score

Widget grading

Many interactive widgets can be graded individually.

Examples include:

• Text input
• Number input
• Multiple choice
• Checkbox
• Dropdown
• Hotspot (participation only: any tap counts as correct)
• Find in picture
• Fill blank
• Matching
• Sorting list
• Drop bucket
• Text highlight
• Interactive rain
• Container

Most of these — every input and interactive widget, plus the spin wheel — arrive graded by default the moment you add them, so a fresh question counts toward the score without an extra step. Turn grading off on any widget you're using for practice, acknowledgement, or support. Flashcards, media, plain text, buttons, and containers start ungraded, since they aren't answers on their own. So do hotspots: a hotspot has no wrong answer (any tap counts as correct), so grading one only hands out points. See Hotspot. To grade whether learners can find things in a picture, use Find in Picture, which starts graded and knows which regions are right.

A widget with grading off adds nothing to any score: not the live Score widget, not the lesson's ending, and not the grade you see for the student. Learners can still answer it, and their answer is still saved. It also still knows whether that answer was right, so On Correct actions and branching on it keep working. You don't need to change its points first.

Grading is set in the Scoring section of each widget's ⚙ Configure dialog. Each graded widget has:

• Scoring — an On/Off switch at the top of the section
• Points — how much the answer is worth
• Group — an optional grading group

A widget with Scoring on counts toward both the scene score and the lesson score. Its name (edited at the top of the properties panel) is how it's labelled in results and conditions.

Some activity widgets have choices that ask students nothing to check: a Pattern Blocks or Connecting Cubes board set to Explore, a Ten Frame or Fingers set to Show, or a Counting Collection set to Just count. Their list of what students do tags every choice Graded or Not graded, so you can see which ones score before you pick. Pick a Not graded one and the Scoring section stays where it is, marked Not graded, and says which choice would grade it. A Not graded widget counts 0 of 0: it never lowers a score and never holds up a scene.

Container grading

Containers can turn several child widgets into one scored unit.

When a container's scoring is enabled, its child widgets are evaluated inside the container, but they do not contribute separately to scene score or lesson score. The container contributes one result instead.

This matters because it prevents double-counting. A group of three child questions can be treated as one five-point task, one pass/fail station, or one custom grading group result.

Container aggregation decides whether the container is correct:

• All children correct requires every graded child inside the container to be correct. A child with grading turned off doesn't count here, and neither does a long answer or drawing, which nothing can grade yet.
• Minimum total score requires the child widgets to earn at least a chosen number of points.
• Minimum percentage requires the child widgets to earn at least a chosen percentage of their combined possible points.

If the aggregation rule passes, the container earns its configured point value. If it fails, it earns zero.

Once the learner answers any question in a container, a question they skip inside it counts as wrong (zero points). One right answer and two blanks does not pass All children correct. Hidden questions (visibility set to Hidden) are left out of the rule, since the learner can't see them.

Scene score

Scene score answers a local question:

How did the learner do on the graded work that belongs to this scene?

Use scene score when you want each step to evaluate itself independently.

This works well for:

• One-scene checks for understanding
• Pass or retry branching
• Immediate scene-level feedback
• Showing learners whether they completed the current task successfully

Lesson score

Lesson score answers a broader question:

How is the learner doing across the whole lesson?

Use lesson score when you want:

• An overall performance display
• A cumulative progress feel
• A final success scene based on the full lesson
• Multiple scenes to contribute to one score

How the live score is counted

The score a learner sees while they work, and the score conditions check, add up the same way the final grade does:

• Lesson score counts every graded question in the scenes the learner has reached so far, answered or not. A skipped question is a zero, so one right answer out of ten reads 1/10, not 1/1. Scenes the learner hasn't reached yet (another branch, say) aren't counted.
• Scene and group scores read the whole lesson. A group score, or a condition on another scene's score, includes questions on other scenes, not only the scene on screen.
• Each widget's maximum is what it can really earn. A spin wheel is worth its best segment times its number of spins. A per-option conveyor belt is worth its points per option times its correct options. A long answer or a drawing counts 0 of 0, answered or skipped, because nothing grades either yet — its Points never reach any score, so leaving one blank can't cost a learner anything.
• Hidden questions don't count unless the learner answers one after an action shows it.
• Questions behind a "show when" or "hide when" rule count only if the rule let the learner see them. When the attempt ends, Berdee checks each such rule against the learner's answers. If question B appears only when question A is right and the learner got A wrong, B isn't in their grade, and a skipped B costs nothing. If they got A right and skipped B, B is a zero. A question the learner answered always counts. See Visibility and Personalization for which rules this covers.

Custom grading groups

Sometimes you want to grade a subset of widgets together rather than just “this scene” or “the whole lesson.”

That is what grading groups are for.

Use a grading group when:

• One page has practice and assessment widgets mixed together
• Only some questions should count toward a checkpoint
• You want a button to submit only one part of a larger scene
• A review challenge should use its own scoring scope

A Submit button set to a grading group grades and locks every widget in that group — including group widgets on scenes the learner already left — and nothing else. A group question the learner skipped counts as 0, but only if they visited its scene; a group question on a branch they were never sent down, or one hidden from them on the scene they submit from, isn't counted. Afterwards, “is submitted” for that group is true, while the scene it was pressed on is not submitted.

To report each skill on its own, put each question in a group named after its skill and show one Score widget per group on the ending. Step-by-step: Score each skill separately with grading groups.

Require submission vs immediate feedback

Scene settings let you control whether grading should wait for submission.

Use submission-based grading when:

• Learners should complete a set of answers before seeing the outcome
• You want a clear Submit moment
• Feedback should feel like a checkpoint instead of constant interruption

When the scene's Progression is set to Stay, Berdee grades the selected scope and keeps the learner on the current scene. It only moves the learner automatically when you choose a destination — the next scene, a specific scene, or a router — in the scene's Progression settings.

Use more immediate feedback when:

• The activity is exploratory
• Learners benefit from fast correction
• The task feels more like guided practice than formal assessment

How the score widget fits in

The score widget is the learner-facing display for grading progress.

It is most useful when you want learners to feel momentum or understand their result at a glance.

Strong uses:

• Keep a running score visible through a game-like activity
• Show completion progress during practice
• Display a final score on a feedback scene
• Reinforce that a multi-scene challenge is cumulative

The score widget becomes more meaningful when grading rules are simple and intentional. If too many widgets count in inconsistent ways, the score display can confuse rather than motivate.

The Label is a name, not a sentence. Berdee puts the colon in for you, so type Your score and learners read Your score: 8. Typing Your score: gets them Your score:: 8. Same for Points earned, This round, Total — leave the punctuation off and let Berdee join the two halves. More detail: Score Widget.

Useful grading combinations

Scene score + require submission

Use this when you want a scene to feel like a complete checkpoint.

Good fit:

• A five-question review at the end of a lesson
• A short exit ticket
• A scene that should branch to pass or retry

To let a few slips through, set the scene's Mark this scene correct when to Cumulative and choose a Passing %. Step-by-step: Move on when students get four of five right.

Lesson score + a score on a shared layout

Use this when the learner should feel progress building across several scenes. Put a lesson-scope score widget on a layout and apply that layout to each scene, so the running total stays in the same spot throughout.

Good fit:

• A vocabulary game that spans several rounds
• A multi-scene science challenge
• A review sequence where each station contributes to the final result

Custom grading group + support widgets on the same scene

Use this when only part of the scene should count.

Good fit:

• A practice-and-check scene where hints and examples are visible but not graded
• A challenge area that lives beside a support panel
• A bonus round embedded inside a larger scene

Container score + multi-part task

Use this when several widgets are really one task.

Good fit:

• A reading panel with three checks that should count as one activity
• A math problem with multiple entry fields that should produce one result
• A station card where several child widgets determine whether the station passes

A practical grading pattern

For a short assessment flow:

1. Enable grading on the answer widgets
2. Decide which ones count toward scene score or lesson score
3. Place a Submit button where the learner commits answers — set its Submits setting to Current scene for a per-scene checkpoint or Lesson to finalize the whole attempt
4. Set the scene's Progression — Stay, the next scene, a specific scene, or a router
5. Show a score widget where it helps reinforce the outcome

Use a regular button for navigation, retry, reveal, reset, and media actions. Use a Submit button when the learner is sending scored answers for checking.

**An attempt ends when the learner reaches the Closing Scene — that's when the final score is written to the gradebook and their answers are locked. A Submit button set to Submits = Lesson also finalizes the attempt wherever it is pressed, and a lesson still needs one to publish; that gate is on its way out, but until it goes both paths end an attempt. Per-scene Submit buttons (with Submits = Current scene**) are optional and act as checkpoints: they grade and lock the scene — a question the learner skipped counts as 0 out of its points and locks with the rest (one hidden from them at that moment isn't counted) — send a partial save to the server, and can branch the learner based on this scene's pass/fail.

Strong real-world examples

• A five-question warm-up where each scene contributes to a lesson score that appears on the final scene.
• A single challenge scene where only the main question widgets count toward scene score, while hints and supporting widgets do not.
• A review game where a custom grading group tracks only the “bonus round” widgets, and a button submits that group separately.
• A station-style activity where each container owns several child answers and contributes one score item to the lesson total.
• A middle-school reading check where the learner sees a scene score right after submission, then moves to a feedback scene with next-step advice.
• A primary-grade phonics lesson where only the final check counts toward score, while practice prompts and audio support stay ungraded.
• A prize-wheel review scene where a spin wheel awards points per landed segment and (with Require spin to complete scene on) blocks scene completion until the learner has spun.

Common mistakes

• Enabling grading on too many widgets just because you can
• Mixing scene score and lesson score without a clear reason
• Forgetting that a scored container replaces its child widgets in scene and lesson totals
• Using score displays before deciding what should count
• Forgetting to test how grading behaves after submission

Recommended next reads

• Make some questions count more, and one not at all (step-by-step tutorial)
• Score a three-part word problem as one question (step-by-step tutorial: container grading)

• Events, Actions, and Conditions
• Interaction Recipes
• Score Widget
• Container Widget
• Classrooms and Assignment Workflows