Available in: ValueEdge
The AI root cause analysis widget helps you establish correlations between a defined set of problematic work items.
Root cause analysis: Overview
The problematic work items root cause analysis widget helps you define problematic items, track them across releases, and identify the root cause for their problematic behavior.
For example, you can use this widget to track and identify the root cause for the following problematic work items:
- Escaped or high reopen ratio defects
- High cycle time features or epics
Features or stories with more than <x> number of related defects
User stories with the invested hours per number of story points exceeding a certain threshold
The widget provides a standard dashboard graph and the AI insights on the right pane.
Use the AI insights to understand what may be causing the work items to be problematic, for example, based on how they correlate with certain properties of the problematic items. For details on the widget insights, see ALM Octane insights.
Root cause analysis: Configuration
Configure the widget as follows:
|Item type||Select the type of work items that you want to track.|
|Problematic work items||
Use this field to define the work items that demonstrate a pattern of undesired behavior which you want to improve.
The Problematic work items field acts in the following ways:
The widget displays the number of problematic work items according to your definition and within a specified time frame.
By default, the widget tracks escaped defects – the defects that have no linked runs and thus, you may assume, have escaped from any formal testing activities. Define any other filter of defects that you consider problematic. For example, you can track defects with a high reopen ratio, or regression defects.
Tip: Change the widget title to reflect the problematic work items' definition.
Correlations allow you to discover patterns in the attribute values that may explain the work items’ problematic behavior.
Correlation fields are specific item attributes that you suspect may be connected to problematic work items.
The AI analytics measures the correlation levels between the specified attributes and the problematic items.
For example, by selecting the Application module as a correlation field, you may identify how correlated problematic work items are with certain application modules. If the correlation score is high, it may indicate that the items’ problematic behavior is rooted in a specific application module, and signal that your testing strategy does not cover the given area well enough.
You can select up to 20 correlation fields. 20 correlation fields are predefined. For example: Application module, Author, Detected by, QA owner, Team, and other.
Edit the predefined fields or select others, as necessary.
|Release, sprint, and milestone||Select the scope in which you want to track problematic work items.|
|Additional filters||Define any other filter to narrow down the scope of problematic work items. For example, you can filter the problematic work items per Team, Program, Environment, or other item properties.|
On the Display tab, configure the widget display options, as necessary. By default, the widget x-axis displays the target release, and the y-axis displays the count of problematic work items.