Multi-Team Data: Analytics

This article is part of our series on Reviewing and Analyzing Multi-Team Data. When analyzing Multi-Team data, you will review a number of areas in AgilityHealth®. We've broken each area down for you, into shorter articles.

Analytics

The Analytics Section is helpful for a quick glance at the highest and lowest competencies and consensus from the radar, but can also help you to dig deeper. This section is located beneath the roll-up radar image or you can get there by clicking the analytics icon in the upper right-hand corner of the screen.

Here’s what to look for:

See whether any of the Top 5 or Lowest 5 Competencies are also within the 5 Highest Consensus Competencies. This indicates that not only did the competency have one of the highest or lowest scores, but that there is strong agreement across the teams.

In the analytics table above, Technical Lead - Leadership and Technical Leadership are both in the Top 5 as well as 5 Highest Consensus lists. You can see this on the radar:

See whether any of the Top 5 or Lowest 5 Competencies are within the 5 Lowest Consensus Competencies. This indicates that while the competency score is, on average, one of the highest or lowest across the teams, there is a lot of variation.

In the analytics table above, Roadmap is the Lowest Competency and it is also in the list of 5 Lowest Consensus Competencies. You can see this on the radar:

See how many of the competencies within either the Top 5 or Lowest 5 Competencies fall within the same area of the radar. If three or more are within the same dimension or sub-dimension, this indicates an important area to focus on, either to recognize what is working well or as an opportunity for improvement.

To learn more about the questions that are asked for each competency in the table and what the score represents in terms of maturity, click on the competency in the table to view the details. If the Growth Portal is enabled for your organization, you can click on the Growth Portal button on the Competency Profile to access more information and resources about each competency.

Learn more about related competencies.

For a TeamHealth assessment, you can also check the competencies in the Performance Dimension and the Happiness competency to see how strong the team is in relation to the top drivers for those metrics. Learn how to see top drivers for team performance.

How is Consensus calculated?

The top 5 highest consensus competencies are the competencies with the least variance and the lowest 5 consensus competencies are the competencies with the highest variance.

Our standard calculation for consensus is very simple and highlights outliers. This calculation takes the difference between the highest score and the lowest score to determine consensus.

For example, assume for a given competency the scores were 9, 9, 9, 9, 8, 8, 8, 8, 8, and 2 on a 10 point rating scale. To find the difference between the highest and lowest score we subtract them: 9-2=7. We then divide 7 by the number of points on the rating scale, which in this case, is 10. So, 7 divided by 10 is 0.70 but to make this a percentage, we multiply by 100 and end up with 70% (7/10*100=70%). We then subtract 70% from our possible consensus score of 100% to get 30% (100%-70%=30%). The consensus for the given competency is 30%.

If the questions within a competency use a different scale, such as an 8 point scale, the calculation is adjusted accordingly. Let's assume the scores for a given competency were 7, 7, 5, 4, 2, and 2 on an 8 point rating scale. The highest score minus the lowest score (7 and 2) would be 5. It would then be calculated as 5/8*100=62.5%. To find the consensus percentage, we take 100-62.5 which equals 37.5%, which is rounded up to 38%.

Percent Standard Deviation

This is not widely used by our customers but we also offer a percent standard deviation model for consensus which highlights overall consensus and does not give as much weight for outliers. The formula for this is the regular standard deviation divided by the average result multiplied by 100.

For the result set above, this would result in a percent standard deviation of 25.5% - but it is important to note that with percent standard deviation the lower the score, the more consensus. 0% standard deviation would mean complete agreement, whereas 100% standard deviation would mean no agreement.

How can we help you?