What happened in this attempt
Four reversals occurred before the prompt. The task was then completed with support. These are event records, not conclusions about the person.
See the problem-solving process behind the answer.
Most assessments stop at right or wrong. Noesis also looks at how you began, where you changed course, and which kind of support helped—then turns repeated patterns into a specific strategy you can practise.
Our advantage is not another all-purpose score. It is an explanation of the route behind the result.
From one attempt to one practical training decision.
Take a meeting-schedule problem as an example. Noesis preserves the useful detail between answering the question and deciding what to practise next.
Illustrative taskPlace four meetings into three time slots. Two meetings cannot be adjacent, and one must occur before another.
The learner starts assigning meetings immediately, reverses course four times, then completes the task after a prompt to list the constraints first.
What this problem required
The solver needed to keep several conditions active, identify which ones ruled out options, and order them before assigning meetings.
Choose one useful move
A short exercise can teach the learner to list and rank constraints before committing to a solution.
Try the move in a new context
A later scheduling problem changes the story and layout, then checks whether the learner selects the strategy without the same prompt.
A profile becomes clearer as the evidence grows.
Noesis keeps early observations lightweight, strengthens them when a pattern repeats, and updates them when later work points in a different direction.
Assisted and independent solutions remain separate. Speed is read alongside accuracy and task demands. This preserves the context needed to make the next recommendation useful.
First signal
A single task creates a local observation and gives the system something specific to check next.
Pattern appears
A similar structure produces the same behavior, making the observation more useful for training selection.
Context changes
The story, values, or presentation change while the same strategy remains relevant.
Profile updates
New independent work can strengthen, narrow, or replace the earlier interpretation.
Seven dimensions turn complex performance into a practical training map.
They connect assessment tasks, profile observations, and strategy practice in language users can understand and act on.
Performance categories
Patterns that require comparison across more than one task and occasion.
Processing speed
Compare alternatives efficiently when the task has a genuine time demand.
Numerical reasoning
Recognise a useful quantitative relationship instead of relying only on routine calculation.
Working memory
Keep relevant information available while updating an intermediate result.
Task-demand categories
Descriptions of what a particular problem asks the solver to do.
Pattern recognition
Infer a rule that accounts for a sequence or set of relationships.
Logical reasoning
Use conditions and exclusions without adding an unsupported assumption.
Strategy-training categories
Named actions that can be explained, practised, and checked in another context.
Problem decomposition
Turn the meeting problem into goals, variables, and constraints before assigning slots.
Cognitive flexibility
Notice that repeated enumeration is not progressing and choose a different representation.
The frontier has moved beyond scores.
Current cognitive science asks how confidence is calibrated, how strategies unfold over time, when people should switch approach, and whether learning survives a change of context. These questions directly shape Noesis.
Calibrate confidence, not just correctness.
- Research direction
- Metacognition research treats confidence as an inference about one’s own performance. It can align with accuracy—or diverge from it in systematic ways.
- How it appears in Noesis
- Noesis keeps correctness, confidence, and revision separate, then looks at whether confidence becomes better calibrated as new evidence appears.
Model the route, not only the result.
- Research direction
- Recent computational work uses sequences of decisions and response times to distinguish processes that a final score cannot reveal.
- How it appears in Noesis
- Noesis preserves the order of starts, hints, revisions, checks, and strategy changes. Two identical answers can therefore lead to different feedback.
Know when to persist—and when to switch.
- Research direction
- Cognitive flexibility is not constant switching. It depends on balancing stable goals with a response to changing demands.
- How it appears in Noesis
- Noesis does not reward changing strategy for its own sake. Training asks the learner to recognise whether the current approach is still productive before staying or switching.
Make adaptation explainable. Test learning at a distance.
- Research direction
- Learning analytics and cognitive-training research increasingly separate the rule that changes practice from the evidence that learning carries into a different task.
- How it appears in Noesis
- Every Noesis recommendation can show what changed and why. Practice, a structurally similar problem, and a wider application remain three distinct checks.
A method designed to stay understandable.
Noesis connects repeated task patterns with the next strategy workout, then uses later practice to show how the approach is changing.
- Recommendations show which repeated observations influenced the choice.
- Assisted learning and independent performance remain separate.
- Task practice, near transfer, and broader transfer are tracked as different outcomes.