# Relationships, joins, unions and data source filters

*Tableau Data Analyst Study Guide · Connect To and Transform Data · Last updated October 29, 2026 · https://www.certifyforce.com/study-guide/tableau-data-analyst/connect-to-and-transform-data/relationships-joins-unions-and-data-source-filters*

Combine tables in the right layer (relationships and multi-fact models in the logical layer, joins and unions in the physical layer), validate a relationship model for missing or repeated values, and choose between data source filters and extract filters.

- Certification: Salesforce Tableau Data Analyst
- Exam section: Connect To and Transform Data
- Study guide: https://www.certifyforce.com/study-guide/tableau-data-analyst

**Official exam objective:**

> Prepare data for analysis: Assess data quality (completeness, consistency, accuracy); Perform cleaning operations; Organize data fields into folders; Use multiple data sources (establish relationships, create joins, union tables); Prepare data by using Data Interpreter, pivot, and split; Create extract and data source filters.

## What the exam expects

- Choose a relationship, join or union from what the analysis needs (more columns, more rows, or tables at different levels of detail), including a multi-fact relationship model when tables would form a cycle.
- Build joins with join clauses and calculations, and unions by name, by field order or by wildcard, merging mismatched fields.
- Assess completeness and accuracy of a relationship model with Count fields, View Data and Show Empty Rows, and set Performance Options safely.
- Create data source filters and extract filters, and know which applies where.
- Data Interpreter, pivot, split, data quality checks and Data pane folders are taught in Cleaning data with Data Interpreter, pivot and split; the basics are in the Desktop Foundations units Relationships versus joins in the data model and Join types and unions in the physical layer.

Key terms:
- **Logical layer:** the Relationships canvas, where logical tables stay separate at their own level of detail.
- **Physical layer:** the Join/Union canvas inside a logical table, where tables merge into one flat table.
- **Multi-fact relationship model:** a data model with more than one base table, linked through shared tables (2024.2 and later).
- **Data source filter:** a filter on the data source that every query of it applies.
- **Extract filter:** a filter that limits which rows are pulled into an extract.

## Key facts

| Fact | Detail | Source |
|---|---|---|
| Relationship join type | None to set; Tableau picks per query | [The Tableau Data Model](https://help.tableau.com/current/pro/desktop/en-us/datasource_datamodel.htm) |
| Relationship limits | Same data type; no geographic fields; no cycles | [Relate Your Data](https://help.tableau.com/current/pro/desktop/en-us/relate_tables.htm) |
| Multi-fact models | Version 2024.2 and later | [The Tableau Data Model](https://help.tableau.com/current/pro/desktop/en-us/datasource_datamodel.htm) |
| Performance Options defaults | Many-to-Many; Some Records Match | [Optimize Relationship Queries Using Performance Options](https://help.tableau.com/current/pro/desktop/en-us/datasource_relationships_perfoptions.htm) |
| Dimension-only view | Inner joins; unmatched domain hidden | [Relate Your Data](https://help.tableau.com/current/pro/desktop/en-us/relate_tables.htm) |
| Split fields | Can't be join keys | [Split a Field into Multiple Fields](https://help.tableau.com/current/pro/desktop/en-us/split.htm) |
| Headerless union | Generate field names automatically | [Union Your Data](https://help.tableau.com/current/pro/desktop/en-us/union.htm) |
| Edit Data Source Filters OK | Promotes every listed global filter | [Filter Data from Data Sources](https://help.tableau.com/current/pro/desktop/en-us/filtering_datasource.htm) |
| Extract filters, multi-fact | Applied per table | [Filter Data from Data Sources](https://help.tableau.com/current/pro/desktop/en-us/filtering_datasource.htm) |

## Check yourself

- Why can setting Cardinality to One distort totals?
- When should you choose a join over a relationship?
- Why can a data source filter slow a multi-table data source?
- How are unrelated dimensions from two base tables combined in one view?

## Sources

- [Relate Your Data](https://help.tableau.com/current/pro/desktop/en-us/relate_tables.htm)
- [The Tableau Data Model](https://help.tableau.com/current/pro/desktop/en-us/datasource_datamodel.htm)
- [Join Your Data](https://help.tableau.com/current/pro/desktop/en-us/joining_tables.htm)
- [Union Your Data](https://help.tableau.com/current/pro/desktop/en-us/union.htm)
- [Filter Data from Data Sources](https://help.tableau.com/current/pro/desktop/en-us/filtering_datasource.htm)
- [Optimize Relationship Queries Using Performance Options](https://help.tableau.com/current/pro/desktop/en-us/datasource_relationships_perfoptions.htm)
- [About Multi-fact Relationship Data Models](https://help.tableau.com/current/pro/desktop/en-us/datasource_mfr_multiple_base_tables.htm)
- [Extract Your Data](https://help.tableau.com/current/pro/desktop/en-us/extracting_data.htm)
- [Split a Field into Multiple Fields](https://help.tableau.com/current/pro/desktop/en-us/split.htm)

## About the full unit

The "Core concepts", "Exam traps" and "Worked scenario" sections of this unit, its 7 linked practice questions, and unlimited timed mock exams are included with the Tableau Data Analyst pack (USD $19): https://www.certifyforce.com/practice-exams