💡 Important to note: Maki assessment scores are indicators and insights designed to support recruitment decisions. Each recruitment team has its own expectations about scores. Scores never automatically qualify or disqualify candidates — even if a score is low, recruiters may override it and advance a candidate to the next stage based on other factors or their own appreciation of the candidate's performance. Your test results are only one part of the overall recruitment decision, which is always made by the recruiter.
1. AI-Scored Activities
Some activities, such as our open-ended language proficiency assessments or structured interviews assessing soft skills, are scored using AI models.
The evaluation criteria depend on the type of activity:
Language assessments
These are scored against the CEFR proficiency framework. Depending on the activity, this may include:
task achievement
coherence and cohesion
lexical range
grammatical range
pronunciation (for spoken responses)
Structured interviews and soft skill assessments
These are evaluated using behaviourally anchored rating scales (BARS). This means that while scoring is supported by AI, scoring is highly structured, transparent, and auditable — allowing recruiters to identify and review how a candidate has objectively performed against predefined criteria.
Each soft skill is composed of multiple sub-dimensions, and each sub-dimension is evaluated using its own scoring grid. The sub-dimension scores are then combined to produce the overall skill score.
Example
Soft Skill: Teamwork
Sub-Dimensions of Teamwork: Collaboration; Supportiveness; Communication
🧠 Example — "Describe a time you worked effectively in a team"
Score | What this level looks like |
5 — Excellent | Gives a specific example, clearly explains their role, describes the outcome, and reflects on what they learned. |
3 — Adequate | Gives a general example of teamwork but lacks specific details about their own contribution or the result. |
1 — Limited | Gives a vague or off-topic answer with no clear example of teamwork. |
The example above is simplified to help illustrate the concept. In our actual assessments, BARS scales are more detailed and include more nuanced behavioural descriptions for each rating level. Each score level is anchored to a clear description of what that performance looks like, so scoring is consistent and transparent for every candidate.
As each activity type has its own evaluation criteria, your final score reflects how well your answers align with the criteria defined for that activity.
2. Deterministic Scoring
Some activities use predefined, deterministic scoring methods. This means scores are calculated automatically based on fixed rules, with no human or AI interpretation involved. While the exact approach may vary by assessment, there are two common patterns:
a. Some assessments give the most points to the best answer, fewer points to the next best option, and no points for the rest.
This applies to both single and multiple choice questions, where answers carry different point values depending on how well they reflect the ideal response.
Single choice example:
Answer 1: 3 points
Answer 2: 2 points
Answer 3: 0 points
If you select Answer 2: Score = 2/3 ≈ 0.67
Multiple choice example:
The score per question is calculated as: sum of points for selected answers ÷ sum of points for all positive answers. In some cases, negative points may apply for incorrect selections, though your score for any question cannot go below 0%.
Answer 1: +3
Answer 2: +3
Answer 3: +2
Answer 4: −1
If you select 2 and 3: 5/8 = 62.5%
Your final score = total points earned across all questions ÷ total number of questions
b. Other assessments simply mark answers as correct or incorrect.
This applies to single choice, numerical, and true/false questions, as well as some multiple choice formats.
For single choice, numerical, and true/false questions: each correct answer earns 1 point; each incorrect answer earns 0. Your score = total points earned ÷ total number of questions.
For multiple choice questions, you gain points for correct selections and may lose points for incorrect ones:
Score per question = correct answers selected ÷ total correct answers − (wrong answers selected × 0.5)
Your score for any question cannot go below 0.
Example (where Answer 1 and 2 are correct, Answer 3 is incorrect):
Select 1 + 3: 1/2 − 0.5 = 0
Select 1 only: 1/2 = 0.5
Select 1 + 2: 2/2 = 1
Select 3 only: 0/2 − 0.5 = 0
Your final score = total points earned ÷ total number of questions
💡 Need additional help?
If you have any questions about this article, please contact us at [email protected]. We are here to help.
Thank you for being part of our Maki community!