How Online Assessment Interviews Are Scored: Rubric by Level (2026)

A online assessment interview is scored on 3 dimensions: correctness (tests passed) (50%), time and space complexity (25%), code quality (25%). Each is rated 1 to 4 against anchors that rise with the level. At mid level, an outstanding score on correctness (tests passed) means: "Passes every test on every problem, including the largest inputs."

The dimensions and their weights

DimensionWeightSolid (3) at mid level
Correctness (tests passed)50%Solves most problems fully; one fails a few hidden edge or limit cases.
Time and space complexity25%The expected complexity on most problems, with one slower than it should be.
Code quality25%Clear structure and names, idiomatic, edge cases handled in one place.

What changes between levels

On correctness (tests passed), the heaviest dimension, an outstanding mid-level answer "passes every test on every problem, including the largest inputs", while an outstanding senior answer "passes every test on every problem, including the largest inputs". The same answer can score 4 at mid level and 2 at senior: senior candidates are expected to raise things unprompted that mid-level candidates may be asked about.

Every score, at every level

Correctness (tests passed) (50%)

Level1 · Poor2 · Borderline3 · Solid4 · Outstanding
Intern / new gradPasses few or none of the tests on any problem.Passes the sample cases on one problem but fails most hidden tests.Passes all tests on one problem and some on the other.Passes every test on every problem.
JuniorPasses few or none of the tests on any problem.Passes the samples but fails most hidden tests, often on edge cases.Passes every test on most problems; misses a few edge cases on one.Passes every test on every problem, edge cases included.
Mid-levelPasses few or none of the tests on any problem.Solves at most one problem fully; the others fail most hidden tests.Solves most problems fully; one fails a few hidden edge or limit cases.Passes every test on every problem, including the largest inputs.
SeniorPasses few or none of the tests on any problem.Solves at most one problem fully; the harder ones fail most hidden tests.Solves most problems fully, including the hardest, with one or two hidden failures.Passes every test on every problem, including the largest inputs.
StaffPasses few or none of the tests on any problem.Solves at most one problem fully; the harder ones fail most hidden tests.Solves most problems fully, including the hardest, with one or two hidden failures.Passes every test on every problem, including the largest inputs.

Time and space complexity (25%)

Level1 · Poor2 · Borderline3 · Solid4 · Outstanding
Intern / new gradBrute force that cannot finish on the stated limits, on every problem attempted.Works on small inputs but is a step slower than the limits need.Meets the limits on most problems with a standard approach.Meets the limits everywhere with the expected approach and no wasted work.
JuniorBrute force that cannot finish on the stated limits.A step slower than the limits need on most problems.The expected time complexity on most problems; space is reasonable.The expected time and space complexity on every problem.
Mid-levelBrute force that cannot finish on the stated limits.A step slower than the limits need, such as quadratic where linear is expected.The expected complexity on most problems, with one slower than it should be.The expected or better complexity on every problem, with no redundant passes or copies.
SeniorBrute force that cannot finish on the stated limits.Slower than the limits need on the harder problems.The expected complexity on every problem; space could be tighter on one.Optimal time and space on every problem, with the data structure that makes it so.
StaffBrute force that cannot finish on the stated limits.Slower than the limits need on the harder problems.The expected complexity on every problem; space could be tighter on one.Optimal time and space on every problem, with the data structure that makes it so.

Code quality (25%)

Level1 · Poor2 · Borderline3 · Solid4 · Outstanding
Intern / new gradHard to follow: one long block, meaningless names, dead code.Runs, but names and structure make it hard to read.Readable names and a clear flow; some repetition.Clear, idiomatic code a reviewer could read at a glance.
JuniorHard to follow: one long block, meaningless names, dead code.Readable in parts; repeated logic and unclear names elsewhere.Clear names and flow, idiomatic for the language, little repetition.Clean, idiomatic code with small helpers where they help and edge cases handled visibly.
Mid-levelHard to follow: one long block, meaningless names, dead code.Works but is tangled: repeated logic, unclear names, leftover debugging.Clear structure and names, idiomatic, edge cases handled in one place.Production-quality: idiomatic, well factored, nothing left over.
SeniorHard to follow: one long block, meaningless names, dead code.Works but is tangled or not idiomatic for the language.Clear structure and names, idiomatic, edge cases handled in one place.Production-quality: idiomatic, well factored, nothing left over.
StaffHard to follow: one long block, meaningless names, dead code.Works but is tangled or not idiomatic for the language.Clear structure and names, idiomatic, edge cases handled in one place.Production-quality: idiomatic, well factored, nothing left over.

Get scored against this rubric

Thita's AI interviewer runs a live online assessment round and scores you against exactly this rubric, dimension by dimension, for the level you choose.

Try a timed online assessment

Where this rubric comes from

These rubrics are researched: they follow the scoring methods big-tech companies describe publicly (behaviourally anchored ratings, bar raisers, leadership principles, named interview signals) and how their rounds are commonly run. No company publishes its actual scoring sheet, so the level-by-level wording is our own synthesis. It is the same rubric our AI interviewer scores every mock against.

Questions, answered

How is a online assessment interview scored?
A online assessment interview is scored on 3 dimensions: correctness (tests passed) (50%), time and space complexity (25%), code quality (25%). Each is rated 1 to 4 against anchors that rise with the level. At mid level, an outstanding score on correctness (tests passed) means: "Passes every test on every problem, including the largest inputs."
What is the difference between a mid-level and a senior online assessment answer?
On correctness (tests passed), the heaviest dimension, an outstanding mid-level answer "passes every test on every problem, including the largest inputs", while an outstanding senior answer "passes every test on every problem, including the largest inputs". The same answer can score 4 at mid level and 2 at senior: senior candidates are expected to raise things unprompted that mid-level candidates may be asked about.
What does a score of 3 mean?
Scores are relative to the level you are interviewing for. A 3 means solid for that level: the bar is met without help on the key points. A 4 means the candidate also raised the important points unprompted. A 2 usually means the interviewer had to lead the candidate to them.
Is this the official rubric used by Google, Meta or Amazon?
These rubrics are researched: they follow the scoring methods big-tech companies describe publicly (behaviourally anchored ratings, bar raisers, leadership principles, named interview signals) and how their rounds are commonly run. No company publishes its actual scoring sheet, so the level-by-level wording is our own synthesis. It is the same rubric our AI interviewer scores every mock against.