Participant Score

[adonahue]

As a user, I want some context and meaning for the individual metrics I get after a meeting, so I understand why it matters and what I should do.

The design goal for this story is:

• Provide a "metrics summary" to the user, that it provides meaning, interpretation, and a behavioral target, so that they are not confused or overwhelmed by all the individual measurements.

Acceptance Criteria

• [x] Short discussion w. @adonahue and @jaedoucette about design possibilities / limitations, and what we can say in good faith.
• Can / should we put a target?
• Can / should we display the breakout of metrics that went into the calculation?
• What is the right score representation - out of 10? 100? as a percentage?
• Other considerations.
• [x] The formula for computing the score ( @jaedoucette )
• [x] A metric at the top of the dashboard that displays a single "participation" score for the meeting. This score represents how well a person communicated with their teammates. Use the visual https://www.amcharts.com/demos/bullet-chart/
• [x] Text on the card back: "Your participation score is a synthesis of how you interacted with other meeting participants. It represents how well you collaborated, and your effectiveness as a communicator."
• [x] Move the timeline to the bottom of the metrics.

[brecriffs]

Seems like more thought/discussion is needed before estimating.

[adonahue]

Possible Visual Displays (with some modifications) https://www.amcharts.com/demos/radial-histogram/
https://www.amcharts.com/demos/angular-gauge/ https://www.amcharts.com/demos/bullet-chart/

[adonahue]

@jaedoucette - can you weigh in on the description text (in story above) and make any suggestions you have about how to explain / describe the participant score? Thx!

[jaedoucette]

I suggest:

"Your participation score is a summary of how effectively you interacted with other meeting participants. This is a beta feature, so if you disagree with our system's assessment, please let us know!"

Ideally, we'd also want to allow the user to drag the bar to the setting they think is most accurate, and then record both the value we showed and the value they suggested in a database. We could then construct a model that would give ratings that more accurately reflected a typical user's self-assessment.

[adonahue]

Thanks @jaedoucette . Is there anything we can say to the user about what makes a good score and how they can improve it? E.g. turn-taking, influence and affirmations will all increase your score.

I think it would help people understand the meaning, and also take positive action.

[jaedoucette]

@adonahue I think that from a long term perspective, it's better not to ask people to change their behaviors, but instead to tell us what they think about the score. The individual components of the score are sound, but right now the score itself may not convey a lot of information about the best way to behave. If we get some data from users, then we can do a simple regression and find proper weights for the 4 components. With proper weights, we could then make more meaningful suggestions to the user about how they can improve.