Safeguarding data quality for trusted insights

Quality responses to study questions are key to conducting successful analysis. As the saying goes, garbage in, garbage out. Almost nothing is more frustrating for researchers than observing inconsistent or unreliable responses after fielding a study. Poor responses to research questions leads to many problems during data analysis including unreliability, reduced statistical power and bad fit of analytic models.

Branded conducts frontline quality assurances to ensure our audience provides quality insights in the studies they complete. But it is imperative that researchers do their part and design research that elicits quality insights from our users.

Branded is dedicated to ensuring quality across our audience. Our optimized AI provides multi-layered quality assurances across all user touchpoints. Our insistence on quality at every user interaction provides you with the data inputs and insights needed to make confident and critical decisions about your products, services and more. 

We use machine learning to predict both poor behavior and unsubstantiated identities of each of the 15k+ new users who register with us every day. Utilizing over 30 data points, continually updated and optimized, we are able to identify, score, classify and remove new users at a 95%+ rate of accuracy.

Branded’s unique and comprehensive scoring and classification system is developed around our total audience, determining whether users are cleared for full participation, require further quality checks and vetting or should be removed from our audience entirely. These frontline assurances ensure data integrity before users enter any study with us. 

In addition to Branded’s commitment to quality, researchers must do their part to design studies that encourage our audience to provide quality responses in the studies they complete. Developing and designing a study may seem easy, but the reality is anything but simple.      

Follow our top 5 tips when designing research to ensure quality insights: 

1. Keep it short and sweet: Limit study length and ask only the key questions you really need to support your research objectives and analysis plan. Break up your research questions across multiple pages when programming to make your study easy to digest. 
2. Be Direct: Use simple and direct language to prevent confusion and frustration from users. Ensure question language is specific to avoid bias and misinterpretation. 
3. Entertain our Audience: Incorporate gamification into your study design whenever possible to keep users engaged and excited about participating in your research. From our user experience research, we know that over two-thirds of our users enjoy playing games on their mobiles and computers. 
4. Pay Attention to Device Preferences: Our audience completes studies across multiple devices, with 42 percent expressing a preference for taking studies on mobile devices and 33 percent expressing a preference for laptops. Preview your study on mobile and desktop to verify that users will have a consistent experience across devices. 
5. Test, Test, Test: Before launching your study, gather as much input as possible from a diverse group of people. Language and an experience that may resonate with researchers may not always resonate with the average person completing a study