ETL Testing ETL Data Transformation Rules 2

Question 1

When testing ETL aggregation transformations, which validation is MOST critical?

Accepted Answer

Verifying that SUM, COUNT, and AVG results match independently calculated totals

Answer

Checking that source file timestamps are preserved

Answer

Confirming that all indexes were rebuilt

Answer

Validating that column names match the source

Question 2

What does 'data standardization' mean in the ETL transformation context?

Accepted Answer

Converting data to a consistent format, such as making all date formats YYYY-MM-DD

Answer

Compressing all data to a standard file size

Answer

Removing all special characters from text fields

Answer

Limiting all numeric fields to two decimal places

Question 3

In slowly changing dimension (SCD) Type 2 handling, what does the ETL transformation do when a source record changes?

Accepted Answer

Inserts a new row with the changed values and sets an expiry date on the old row

Answer

Overwrites the existing target record

Answer

Deletes the old row and inserts a new one

Answer

Flags the old row as an error

Question 4

Which ETL transformation technique is used to combine first name and last name fields into a single full name column?

Accepted Answer

Concatenation

Answer

Splitting

Answer

Pivoting

Answer

Aggregation

Question 5

When testing a NULL-handling transformation rule that substitutes 0 for null numeric values, what test case is ESSENTIAL?

Accepted Answer

Send a record with a null numeric field and verify the target contains 0

Answer

Send a record with value 0 and verify it stays 0

Answer

Send a record with a negative number and verify it becomes null

Answer

Send a record with a string and verify it converts to null

Question 6

What is 'data splitting' in ETL transformation?

Accepted Answer

Separating a single source field into multiple target fields, such as splitting a full address into street, city, and zip

Answer

Dividing a large ETL job into parallel threads

Answer

Partitioning a target table for performance

Answer

Splitting source files into chunks for incremental loading