Skip to main content

Run Statistics

After you have selected the input and configuration, you are ready to run Statistics.

A successful Statistics run produces measures that are accurate enough to support review, reporting, or decision-making.

Typical run steps​

A typical run looks like this:

  1. open Statistics
  2. select the input file or dataset
  3. choose the Statistics configuration
  4. review the fields, groups, and measures
  5. start the Run
  6. inspect the summary output
  7. investigate surprising metrics before using the result downstream

What to check before you run​

Before starting the Run, verify:

1. The input is the correct data​

Confirm that the file is the intended source for the summary.

Check items such as:

  • reporting period
  • source system export
  • business unit or operating context
  • whether the data has already been filtered, joined, cleaned, or deduplicated

2. The configuration matches the question​

Make sure the selected configuration summarizes the right fields and groups.

This matters when similar summaries exist for different teams, periods, or reporting packages.

3. You know what results should be plausible​

Before the run begins, decide what you expect.

Examples:

  • Record counts should be close to a prior Run
  • total amounts should fall within a known range
  • each expected group should appear
  • unusually high or low values should be reviewed

Review the result after the run​

After processing finishes, inspect the summary deliberately.

Review:

  • totals and counts
  • averages, minimums, and maximums
  • grouped results
  • blank or zero handling
  • outliers or unexpected values
  • comparisons to prior runs or known control totals, when available

If the numbers look wrong​

Investigate before sharing the result.

Common causes include:

  • the wrong input file was selected
  • the configuration used the wrong field
  • duplicate records inflated totals
  • values were formatted as text
  • Records needed to be filtered first
  • the grouping field contained inconsistent values

When to rerun after preparation​

Rerun Statistics after upstream data preparation if:

  • values were cleaned or standardized
  • duplicates were removed
  • records were filtered to the correct scope
  • fields were joined from another dataset
  • source data errors were corrected

Final checkpoint before downstream use​

Before sharing the summary, confirm:

  • the correct input was used
  • the intended configuration was run
  • metrics were reviewed against expectations
  • surprising values were investigated
  • the summary answers the business question