Rep Data's In-Survey Quality Tool, ReDem, Helps Bauer Media Save Time, Standardize Data Cleaning, and Improve Survey Data Quality

What were your biggest challenges in ensuring data quality?

When we share survey results with our colleagues, we are often asked, "Who are these respondents? Where do they come from?" We need to be absolutely confident that every response comes from real readers.

To guarantee this, we used to invest a significant amount of time manually reviewing every single interview, checking for suspicious response times, assessing the plausibility of answers, and reading open-ended responses. The sheer amount of time required was by far our biggest challenge.

Why did you decide to implement ReDem after testing it?

We were impressed by its user-friendliness. It requires almost no extra time to upload datasets, select questions for review, and receive results.

Beyond that, the tool seamlessly integrated into our existing manual processes. The software analyzes the same quality criteria we previously checked manually, giving us immediate confidence in the results while significantly reducing the time required for data cleaning.

How do you use the quality tool in your processes today?

We use ReDem for all our surveys. Since most of our survey tools don't support live integration, we manually upload datasets to ReDem after fieldwork and export the cleaned data.

Responses with a ReDem Quality Score below 40 are automatically excluded. For scores between 40 and 59, we conduct a final manual review.

Which feature benefits you the most?

In projects where we can use live integration, we highly value its benefits. Live integration instantly filters out poor-quality responses and enables real-time re-recruitment, which is particularly valuable for complex studies.

We also highly value the Time Score check. In addition to analyzing overall survey completion time, it detects irregular response patterns for individual questions. Manually comparing response times per question would be highly time-consuming, making this an efficient solution for identifying bots.

Has ReDem helped reduce the time spent on data cleaning?

Yes, the time savings are significant. ReDem automates a substantial part of our data cleaning process, reducing manual effort without compromising our quality standards.

A great example is our tracking study, where we survey magazine readers six times a year about their product satisfaction. Each wave includes around 1,000 respondents from various panels. For this study alone, we save an estimated ten hours.

How has data quality improved compared to manual review?

We've observed that the tool detects cases that would have been missed in a purely manual review. In the tracking study mentioned earlier, for example, ReDem identified 10 additional low-quality cases that we wouldn't have found through manual checks alone.

The improvement in data quality is particularly noticeable with live integration. In another study where we were able to use this feature, we saw a clear enhancement in data quality even before fieldwork ended. Participants received products at home after a preliminary survey. We saw a significant reduction in duplicate addresses and incorrect postal returns compared to when ReDem was not in place.

Beyond time savings and improved data quality, have there been any other benefits?

Yes, for example, greater standardization and objectivity in data cleaning. In our tracking study, different team members handle data cleaning in different waves. With ReDem, we avoid subjective decisions and ensure that each wave is cleaned using the same criteria.

Additionally, the tool's objective cleaning criteria allow us to assess data quality across different samples, making it especially valuable when comparing panel providers.