Risk UI components - Linear scale for risk scores #278
jumpertrunks
started this conversation in
Components and patterns
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Background information
Do you think this component or pattern would be useful for other teams and services in MOJ?
Yes, for probation and prison teams needing to display risk information.
Name of product(s) this component or pattern relates to:
All HMPPS digital products and services that include risk information
Team developing it is Assess Risks and Needs in the Probation.
What the product(s) do:
Shows a scale as a horizontal line and gives a visual indication of where an individual’s risk score is positioned on that scale. The scale may be shown numerically, either as percentages or as absolute numbers, or it may use descriptive terms such as ‘low’, ‘medium’ and ‘high.’ Whether the scale is numeric or text-based will vary depending on the output of the different actuarial risk tools that generate the scores.
The scale is divided into coloured zones that represent groupings of scores, for example: low, medium and high. The scale is a continuum, with the lowest score at the left and the highest at the right. A call-out component is used to emphasise the score for the individual. It helps locate the point at which they sit on the scale and shows where that point is in relation to the boundaries between the score groups.
Who the users are:
What the user need is for this component/pattern:
help users assess a risk score for an individual, generated from an actuarial tool such as Risk of Serious Recidivism (RSR)
make it easier for users to rapidly understand risk information without navigating away from their immediate task
support third parties who provide services in their assessment of risk (for example, interventions providers receiving a new person on probation)
What already exists
Similar published components or patterns
N/A
GitHub discussions on similar topics
N/A
Have you tried using what already exists?
There is currently no single consistent HMPPS visual language for communicating risk in digital services.
When to use
Use this component/pattern to help users…
Types of scale and call-out for each risk tool:
Risk of Serious Recidivism (RSR):
Numeric scale with coloured line segments. Call-out uses text + numbers.
OASys Sexual Reoffending Predictor / Contact (OSP/C):
No numbers or text on the scale. Use coloured line segments only. Call-out uses text.
OASys Sexual Reoffending Predictor / Images (OSP/I):
No numbers or text on the scale. Use coloured line segments only. Call-out uses text.
How to use
When using this component or pattern, what should be considered when it comes to:
Establishing the scale
Use the lowest and the highest possible scores for the scale’s start and end points. Some tools will generate a score on a 0-100 scale (absolute values or percentages). Others may start above 0 and/or finish below 100 (percentages or absolute values). The Data Science team can advise. Do not use a scale that ‘squashes’ the majority of results at one end because its span is too wide.
For the call-outs, if the output is numeric, you should also use the text for the description of the score in the call-out, for example: ‘low’, ‘medium’ or ‘high’. The text is shown above the numeric value in the call-out. For text-only outputs, use the descriptions only.
Writing content
Use the definitions of the risk tools and their scores that have been written by the Assess Risks and Needs project. These definitions have been checked with the Data Science team and include advice and instructions on how the users should interpret and utilise the scores.
making sure it’s accessible?
Colour contrast of text meets AA standard
Colour is not the only indicator for a level, text is used too
Implementing in Figma?
A library of Risk components is available in the MoJ Figma account called Risk Components
It can also be viewed here
implementing the code?
TBC
using it in context with other components and as part of a service?
Scales used in the RSR results page
![03a Scores](https://user-images.githubusercontent.com/65774482/142231398-6ed8bfa0-1d5d-42c2-ba09-b6b565bf5dc4.jpg)
when to use variations? (if applicable)
Scale with numbers v scale with no numbers
This will depend on the user need; very quick at a glance information for simple scales to show a risk level might use a scale with no numeric boundary indicators (they may not have them anyway eg OSP scores). Numeric boundaries when a more detailed view is required and the user needs the full context of a score.
Things we don’t know enough about
Beta Was this translation helpful? Give feedback.
All reactions