Score normalization
In Smart Assessment Engine normalization maps the actual scores to a standardized scale, typically from 0 through 100, enabling for a unified evaluation framework.
Normalization approach overview
In Smart Assessment Engine normalization adjusts individual assessment scores to a common scale, enabling comparison across different assessments. Normalization of assessment scores can be achieved through two approaches: Maximizing values, where higher scores are desirable, and minimizing values, where lower scores are preferred.
| To maximize the value | Maximize the value when higher values are better. The following formula is used to get the normalized score: `Normalized Score (High) = ((raw score - minimum Score) / (maximum score - minimum score)) * (scale factor maximum - scale factor minimum) + scale factor minimum` |
| To minimize the value | Minimize the value when lower values are better. The following formula is used to get the normalized score: `Normalized Score (High) = (1 - (raw score - minimum Score) / (maximum score - minimum score)) * (scale factor maximum - scale factor minimum) + scale factor minimum` |
Where,
- Raw score = Actual score calculated
- Minimum score = The minimum value of the current scale.
- Maximum score = The maximum value of the current scale.
- Scale factor maximum = The minimum value of the target scale.
- Scale factor minimum = The maximum value of the target scale.
- Scale definition = Specifies the scoring direction: 'high' for higher scores being better or 'low' for lower scores being better.
- Relative Scaling = When enabled the normalized score is calculated relative to the maximum value.
Note: This approach is applicable only when the linear normalization strategy is used. The formulas may vary if a different strategy is selected.