Skip to content
LakeBench
ProblemsCommunityPricing
Sign inStart practicing
Back
  1. Home
  2. Interview prep
  3. Dimension vs metric in data quality

Data quality · Extra High-Value

Dimension vs metric in data quality

Mediumquality-16
quality dimensionsmetricsnull ratemeasurement

Question

In data quality, what is the difference between a dimension and a metric?

Solution

Here "dimension" does not mean a star-schema dimension table. In quality conversations:

Quality dimensions are *categories of goodness* you care about: accuracy, completeness, consistency, timeliness, validity, uniqueness.

Quality metrics (or measures) are *numeric signals* you compute to score those dimensions over time.

dimension: completeness
metrics:   null_rate(email), missing_partition_count, required_field_fill_%

dimension: freshness
metrics:   lag_minutes = now - max(loaded_at)

dimension: uniqueness
metrics:   duplicate_pk_count, duplicate_pk_rate

Why the split matters

Saying "we care about completeness" is a goal. Saying "alert if email null_rate > 2%" is an operable metric with a threshold.

Tiny example

  • Dimension: validity
  • Metric: pct_rows_with_invalid_status
  • Threshold: warn at 0.1%, error at 1%

Interview tip: "Dimensions = what kind of quality; metrics = how we measure it." Map one dimension to two concrete metrics.

PreviousNext