Microsoft Certified: Power BI Data Analyst AssociateModel the dataMedium

A data modeler is integrating data from multiple regional databases into a single Power BI model. The regional databases have slightly different column names for the same customer identifier (e.g., 'CustID', 'CustomerID', 'Client_ID'). To ensure a consistent and unified customer dimension table, what is the most effective ETL step the modeler should perform?

  1. ARename and standardize the customer ID columns in Power Query Editor.
  2. BUse the MERGE function in Power Query to combine the customer tables.
  3. CCreate a new calculated column in Power BI Desktop to standardize the ID.
  4. DEstablish cross-table relationships in Power BI Desktop using multiple columns.
Show answer & explanation

Correct answer: A. Rename and standardize the customer ID columns in Power Query Editor.

Renaming and standardizing columns in Power Query Editor is a crucial ETL step for data unification, ensuring consistent naming conventions before loading data into the model, which prevents issues with relationships and calculations.

Why the other options are wrong

  • B. MERGE combines tables, but it won't standardize column names across different sources before merging or appending.
  • C. Calculated columns are for DAX calculations after data load, not for standardizing source column names during ETL.
  • D. Establishing relationships in Power BI Desktop assumes consistent column names and won't address the underlying naming discrepancies.

Data Unification (ETL)

Data unification in ETL involves combining disparate data sources into a single, consistent, and coherent dataset, often requiring standardization of formats, naming, and structures.

  • Ensures consistency across multiple data sources.
  • Typically performed in the Power Query Editor (Extract, Transform, Load).
  • Includes renaming columns, changing data types, and handling missing values.
  • Crucial for building a robust and reliable data model.

Memory trick: Unifying data is like assembling a puzzle: each piece needs to fit perfectly.

More Model the data questions