Concept: Data Quality Module
The
Data Quality
module checks your field mapping configuration data to
verify that the data satisfies the usefulness, completeness, and correctness
requirements of People Analytics and Storyteller. The Data
Quality
module checks these data for issues that will negatively impact
your People Analytics installation and application and returns data quality defects that
may need to be addressed before proceeding.The
Data Quality
module currently
checks the field mapping configuration data for the Worker and Hiring pipelines, not
the Skills pipeline.Workday automatically runs the Data Quality module when you install or reinstall People
Analytics, but you can also run it manually. Workday recommends that you manually run
the
Data Quality
module before reinstalling People Analytics to
check that any newly made configuration changes meet all requirements. You can run Data
Quality from the Configure People Analytics
report and view the
results by accessing the People Analytics Data Quality
report.Some defects have a greater negative impact than other defects. Workday assigns
a severity level to each defect:
- Critical. Critical defects can prevent People Analytics from installing successfully. These defects can also severely impact the quality of data and the integrity of analysis results, and can potentially impact constrained security. If encountered, address the critical defects in theConfigure People Analyticsreport or in your underlying data.
- Major. Major defects won't prevent a successful installation, but they can severely impact the quality of data and the integrity of analysis results. Workday recommends that you address these issues in theConfigure People Analyticsreport.
- Moderate. Moderate defects can impact the quality of the KPIs, visualizations, and stories in People Analytics. You may ignore them as their impact on the quality of data is minimal.
The Data Quality module includes these check types:
- Cardinality
- Cross Pipeline Consistent Values
- Cross Pipeline Data Types
- Duplicates
- Illogical Values
- Invalid Values
- Missing Values
- Multiple Hierarchies
- Population
Cardinality
Checks the number of unique values in a particular field. Compares the number to a threshold defined for that field, and checks whether the cardinality for the field is consistent in terms of expected levels within the hierarchy. Workday returns the field that fails the check and shows the cardinality of the field in the Value column.
High cardinality (a lot of unique values) can negatively affect Storyteller engine performance, and it might also result in low quality stories due to a large number of smaller populations.
Cross Pipeline
Consistent Values
Checks the consistency of values across pipelines and compares the two
pipeline-to-pipeline values to a defined threshold. Currently, the threshold is when
no values overlap in the same field in each pipeline.
Workday returns the fields from each pipeline that fail the check and shows the
percentage of inconsistent values from each pipeline in the Value column.
Data Type
Checks whether a particular field in multiple pipelines is the same by inspecting the field type. Workday returns the field that fails the check and shows the data types for the field in the Value column.
Cross pipeline defects can severely impact filtering on the KPI tab by preventing a filter you configure from filtering data in multiple pipelines.
Duplicates
Checks whether there are multiple records of the same worker. Workday returns the Employee ID or Candidate ID and shows the number of multiple records in the Value column.
Duplicate records cause inaccurate headcounts, potentially with exponential impact.
Illogical Values
Checks if a particular field contains values that fail logic rules defined for that field. Example: if the hire date is in the future. Workday returns the field that fails the check and shows the percent of illogical values for that field in the Value column.
Illogical values can lead to inaccurate metrics in topic areas, leading to inaccurate data in
visualizations, KPIs, and stories.
Invalid Values
Checks if a particular field contains values that aren’t in the list of valid values defined for that field. Workday returns the field that fails the check and shows the percent of invalid values for that field in the Value column.
Invalid values can lead to inaccurate metrics in topic areas, leading to inaccurate data in visualizations, KPIs, and stories.
Missing Values
Checks how many values a particular field is missing. Missing values can be blank or NULL. Workday returns the field with missing values and shows the percent of missing values for that field in the Value column.
Missing values might skew the data presented in People Analytics, preventing an accurate reflection of your workforce.
Multiple Hierarchies
Checks whether a child in a hierarchy has parents across multiple hierarchies. Workday returns the field of the child hierarchy with parents across multiple hierarchies and shows the number of multiple hierarchies in the Value column.
A multiple hierarchies defect can result in inaccurate data due to duplication, and it can negatively impact constrained security by potentially displaying to a viewer workers who aren’t in their organization.
Population
Checks the number of unique values in a particular VIBE Index intersection, and compares the number to the minimum headcount requirement defined for VIBE Index intersections. Workday returns the intersection field and the number of intersections that fail the check, and shows the number of records for intersections that don’t have sufficient populations in the Value column.
Population check defects cause Workday to exclude a particular intersection from the VIBE calculation.