Sometimes users notice that a dimension score does not exactly match the average of the statement scores displayed within that dimension. This article explains why that happens and how Comparative Agility calculates dimension scores.
How Dimension Scores Are Calculated
Dimension scores are calculated using all applicable responses across all statements within the dimension.
Responses marked as N/A (Not Applicable) are excluded from the calculation.
Rather than averaging the statement averages, the system calculates the dimension score by combining all applicable responses and then calculating a single average from those responses.
This approach ensures that each individual response contributes equally to the final score.
Why the Dimension Score May Differ from the Average of Statement Scores
The difference becomes noticeable when statements within a dimension have different numbers of applicable responses.
For example:
| Statement | Applicable Responses | Statement Average |
| ———– | ——————– | —————– |
| Statement 1 | 2 | 4.0 |
| Statement 2 | 7 | 5.0 |
| Statement 3 | 7 | 4.86 |
If you average the statement averages, the calculation would be:
(4.0 + 5.0 + 4.86) ÷ 3 = 4.62
However, this method gives equal weight to each statement regardless of how many responses contributed to it.
Comparative Agility instead calculates the dimension score using all applicable responses:
* Statement 1 contributes 2 responses
* Statement 2 contributes 7 responses
* Statement 3 contributes 7 responses
Total applicable responses: 16
If the sum of all applicable responses is 77, the dimension score becomes:
77 ÷ 16 = 4.81
Because this method considers every response individually, the final dimension score may differ from the average of the displayed statement averages.
What Happens with N/A Responses?
N/A responses are always excluded from calculations.
N/A values may appear for different reasons, such as:
* A question was disabled in the survey.
* Responses to a question were removed.
* Respondents selected N/A as their answer.
Regardless of the reason, N/A responses are not included when calculating statement averages or dimension scores.
Why Some Reports Appear to Match
In cases where all statements have the same number of applicable responses, averaging the statement averages and averaging all individual responses will produce the same result.
This is why some reports may appear to match a manual calculation while others do not.
The difference only becomes visible when statements have different numbers of applicable responses due to N/A responses or varying response counts.
Why We Use This Method
Calculating dimension scores from all applicable responses provides a more accurate representation of the data because each response is weighted equally.
This prevents statements with very few responses from having the same influence on the final dimension score as statements with many responses.