Phase 7: Data Checks — Ensuring Data Quality

The final phase before delivering data to the client is a thorough validation check. This process ensures that the data is clean, logical, and accurately reflects the survey’s design. Kalever streamlines this critical step by automating the creation of validation scripts and keeping the resolution process within its collaborative ecosystem.

Scenario: The soft launch data for the ‘Luxury Brands’ survey is gathered, and the raw data file is ready. It’s now time for the data processing team to validate the data before proceeding with the fieldwork.

For a seamless workflow, the Data Processor (DP) must have access to the project’s core information. The Project Manager should ensure the DP is added as a member of the internal project (“Luxury Brands – Internal”).

This grants them access to the Survey Creator, where they can see the survey’s final structure and logic, and to TrackEntry, their primary tool for communicating with the Project Manager and Programmer.

• For instructions on adding team members, see our guide:
  Managing Notifications, Members, and Project Settings

Traditionally, data processors write validation scripts (like SPSS syntax) from scratch—a time-consuming and error-prone process. Kalever’s Data Validations feature automates this entirely.

1. The Data Processor navigates to the Survey Creator within the internal project.
2. They open the Editor Options (cog icon) and select “Data validations”.
3. The system instantly generates a full SPSS syntax script based on the survey’s structure, logic, question types, and valid ranges.

The DP can then download this script and use it to check the raw data file, saving hours of manual work.

• To learn where to find this feature, see our guide:
  Advanced Options: Exports, Imports, and Version Control

This validation script is not a one-time tool. It can and should be used multiple times throughout the data collection process to catch issues early.

Best Practice Workflow:

• After Soft Launch: Run the script on the initial soft launch data to identify and fix any major issues before full launch.
• Ad-Hoc Mid-Field Checks: If any potential problems are flagged during fieldwork, the script can be run on an interim data cut.
• Before Final Data Delivery: Run the script one last time on the complete dataset to perform the definitive, final check.

If the validation script finds an error, the resolution process should be managed within Kalever to maintain a clear audit trail.

Scenario: The SPSS script flags an unexpected value in the data for question S5, indicating a potential programming error.

1. The Data Processor creates a New Entry in the internal project’s TrackEntry board.
2. They describe the issue clearly (e.g., “Data validation script found ’99’ in Q5, which should be a terminated value. Please investigate.”).
3. The programmer is notified, investigates the issue, and communicates their findings or confirms the fix within the same ticket.

This keeps all data-related queries and resolutions documented and tied to the project, ensuring a fully transparent and auditable data cleaning process.

---