Assessment Metric Results Value Calculation

[Kuldeep Kumavat]

Hi,

There is a requirement to configure custom assessments for calculating the T-Shirt size of a Demand.

However, we were not able to figure out the logic/formula used to determine the weighted and the Normalized value of a Metric Result

All our Metrics are of Assessment Type so that the Actual and the Scaled values are same.

The wiki mentions the below about Normalized value (Assessment Results - ServiceNow Wiki😞

Adjusted value that accounts for weights, scale definition, minimum and maximum values, and other factors that impact the metric.

Can anybody help with the logic/formula used to calculated the Weighted and the Normalized Value of a metric.

Regards,

Kuldeep

[anna_scheib]

Hi Kuldeep.   To answer the question, we have a Category score and Metric scores. The sum of weighed metrics for each category gives us the category score. Then the category score is multiplied by a category weight. This weight is different depending on the total number of assessment metrics for a demand. In addition, it doesn't always add up to 100, its rather a relative weight. For instance, if we have a small Size (meaning the scaled result for the Size category is 2), we then multiply this by say 10/80 (where 10 is Size weight and 80 is the sum of all weights), so you get .25 as a normalized value for Size.

Hope this helps.

[jmcquay1 · Accepted solution]

We had a ticket opened with SN and received the following calculation for the normalized value in Metric Results.   I hope somebody finds it useful.

(Input Value - Min value defined in metric) / (Max value defined in metric - Min value defined in metric) * current metric weight / (sum of valid metric weight) * scale_factor defined in metric type

[Kuldeep Kumavat]

This was the closest explanation for the calculation. Hence, marking it correct

[Rohan11]

Hi We too faced similar situation:

Example is below regarding the calculation:

1. Survey Definition -> 1 Category -> 3 metrics (service, staff, timeliness)

For each of them we use the smiley faces very dissatisfied (1) -> very satisfied (5). in the middle we have other values such as dis-sat, neutral, satisfied.

Now if you do a survey where you choose very satisfied (5) for all responses then calculation for normalized value is as follows:

Say for Service Metric (please note scale factor is always 10 unless you changed it)

Input value = 5 (very satisfied)

Min value = 1

Max value = 10

Metric Weight = 10

Normalized value = (Input Value - Min value defined in metric) / (Max value defined in metric - Min value defined in metric) * current metric weight / (sum of valid metric weight) * scale_factor.

So:
(Input Value - Min value defined in metric) / (Max value defined in metric - Min value defined in metric) = (5 – 1) / (10 – 1) = 4/9 = 0.4444

Scale factor = 10.

current metric weight / (sum of valid metric weight) = 10/(10 + 10 + 10) = 10 / 30 = 0.333

Normalised Value = 0.4444 * 0.333 * 10 = 1.481 ~ 1.48.

This means normalised value in such scenario is 1.48 (high values good)

Category Normalised value (if every thing was very satisfied) = 1.48 * 3 = 4.44 

This much I could verify Hope this helps !!