Skip to main content

Regression Testing vs. Sanity Testing: Detailed Explanation with Example

 

Regression testing and sanity testing are both essential software testing techniques, but they serve different purposes in ensuring software stability after modifications.

Regression Testing

Definition: Regression testing is a comprehensive testing approach that ensures recent code changes do not negatively impact the existing functionality of an application. It involves re-running previously executed test cases to verify that the software still works as expected after modifications such as bug fixes, feature additions, or updates.


Key Characteristics:

  • Scope: Covers the entire application.
  • Purpose: Ensures that new changes do not break existing functionality.
  • Execution Time: Time-consuming due to extensive testing.
  • Test Cases: Uses a large set of test cases.
  • Automation: Often automated for efficiency.
  • Depth: In-depth testing of all functionalities.
  • When Used: After major updates, bug fixes, or new features.

Example of Regression Testing

Imagine you have developed an e-commerce website. Initially, you test all functionalities, including:

  • User registration
  • Product search
  • Adding items to the cart
  • Payment processing

Later, you decide to add a new payment method (e.g., cryptocurrency payments). To ensure this addition doesn’t break existing functionalities, you:

  1. Create new test cases for the cryptocurrency payment method.
  2. Re-run all previous test cases to verify that core features (user registration, product search, cart functionality, etc.) still work properly.
  3. If any issues arise, developers fix them before releasing the update.

Sanity Testing

Definition: Sanity testing is a quick and focused testing approach that verifies whether specific changes or bug fixes work correctly before conducting deeper testing. Sanity testing acts as a checkpoint before regression testing, ensuring that the build is stable enough for further testing. If sanity testing fails, regression testing is postponed until the issues are resolved.

 

Key Characteristics:

  • Scope: Focuses on a few critical functionalities.
  • Purpose: Verifies that specific changes work correctly before deeper testing.
  • Execution Time: Quick and limited testing.
  • Test Cases: Uses a minimal set of test cases.
  • Automation: Usually performed manually.
  • Depth: Surface-level testing of key areas.
  • When Used: After minor changes or bug fixes.

Example of Sanity Testing

Consider the same e-commerce website. Suppose a bug was reported where users were unable to apply discount coupons during checkout. The development team fixes the issue, and before running a full regression test, they:

  1. Perform sanity testing by checking whether the discount coupon functionality now works correctly.
  2. Verify that the checkout process is not affected.
  3. If sanity testing passes, they proceed with full regression testing.


Key Differences Between Regression and Sanity Testing

Aspect Regression Testing Sanity Testing
Purpose Ensures that recent changes do not break existing functionality Verifies that specific changes work correctly before deeper testing
Scope Covers the entire application Focuses on a few critical functionalities
Execution Time Time-consuming due to extensive testing Quick and limited testing
Test Cases Uses a large set of test cases Uses a minimal set of test cases
Automation Often automated for efficiency Usually performed manually
Depth In-depth testing of all functionalities Surface-level testing of key areas
When Used After major updates, bug fixes, or new features After minor changes or bug fixes

 

Conclusion

Sanity testing acts as a checkpoint before regression testing, ensuring that the build is stable enough for further testing. If sanity testing fails, regression testing is postponed until the issues are resolved.


Labels:

Comments

Post a Comment

Popular posts from this blog

Understanding Mistakes in Software Development: Errors, Defects, and Bugs

  Every software team uses the words “error,” “defect,” and “bug,” often interchangeably. But there’s real power in knowing exactly what each term means—and when it applies   1. Mistakes by Phase Phase What You Find What It’s Called Requirements & Design A mistake in the design or plan that doesn’t meet what stakeholders want. Defect Coding A coding or logic mistake in source code Error Testing & Execution An observable malfunction occurring during software execution or testing. Bug  🐞 1.1 Defect A defect is any flaw or mismatch in your requirements or design artifacts. It exists before any code runs. Example: You document “Users must enter a 4-digit PIN,” but stakeholders actually needed 6 digits. That spec mismatch is a defect . 1.2 Error An error is a mistake made while coding —a typo, wrong opera...
Post a Comment
Read more

What Is a Feature Flag?

  A feature flag (also known as a feature toggle) is a powerful software development technique that allows developers to enable or disable specific functionality in an application without changing the code or redeploying the software. 🧠 Core Concept Feature flags act like switches embedded in your codebase. They control whether a feature is active or inactive at runtime. This lets teams test, release, or hide features dynamically. 🚀 Benefits Safe Deployments : Deploy code with features turned off, then activate them when ready. A/B Testing : Roll out features to a subset of users to gather feedback. Quick Rollbacks : If something breaks, just flip the flag off—no need to revert code. Continuous Delivery : Decouple feature releases from code deployments for smoother CI/CD pipelines. User Segmentation : Tailor experiences for different user groups. 🧩 Types of Feature Flags Type...
2 comments
Read more

Test Data Design Techniques with Examples

  ✅ 1. Equivalence Partitioning (EP) What it is: Divides input data into partitions of equivalent data from which test cases can be derived. The idea is that if one value in a partition works, all others will too. Example: If a form accepts ages from 18 to 60: Valid partitions: 18–60 Invalid partitions: <18 and >60 Test cases: One value from each partition: 25 (valid), 17 (invalid), 61 (invalid) When to use: When input data can be grouped into ranges or categories. ✅ 2. Boundary Value Analysis (BVA) What it is: Focuses on values at the edges (boundaries) of input ranges, where defects are most likely to occur. Example: For an input field that accepts values from 1 to 100: Test boundaries: 0, 1, 2 and 99, 100, 101 Test cases: 0 (just below), 1 (lower boundary), 2 (just above) 99 (just below), 100 (upper boundary), 101 (just above) When to use: When input fields have defined minimum and maximu...
Post a Comment
Read more