# Data Quality and Anomaly Investigation Workflows

Profile tabular data, isolate duplicates, missingness, referential breaks, outliers, and time-series anomalies, and turn findings into defensible quality evidence.

> Canonical page: https://elysiatools.com/en/hubs/data-quality-anomaly-workflows

- **Keywords:** data quality workflow, duplicate detection, missing data analysis, outlier investigation, time series anomaly detection

## Frequently asked questions

### Does every outlier indicate bad data?

No. Outliers can be valid rare events, so they need business context, not just statistical distance.

### Should duplicates always be deleted automatically?

No. Some repeated rows are legitimate recurring transactions or status snapshots and must be reviewed against keys and timestamps.

### Is interpolation safe for all missing values?

No. Interpolation is appropriate only when the series is ordered and continuity is a defensible assumption.

### Why compare robust means as well as z-scores?

Trimmed and winsorized means show whether a few extreme values are distorting the central tendency enough to change operational decisions.