Data Science

Generate and analyze valid, accurate data

If you want to create research worthy of attention, it needs to be backed by valid, accurate data and a solid methodology. (And it needs to convey a compelling story, but that’s coming up next.)

Let’s walk through the basic steps:

Ask the right questions 

You may think, "How tough can it be to pull together some questions?", but it’s much more nuanced than you might think.

(At Mantis we have a PhD data scientist who reviews every survey, ensuring questions are clear and accurate, and lead to statistically relevant results. And all surveys undergo testing before launch.)


Get enough of the right people 

Once the survey is designed, you need to collect responses. This step of the process was cited as the biggest challenge in survey-based research.

You need either access to a sizeable audience to survey, or budget to access a list. (In some cases, you can partner with someone who has the right list.)

You also need to select the right survey tool to keep surveys as simple and straightforward as possible for participants -- and for you to get the results based on the segments you need.

TIP: Find a partner who has access to the right tool.

Ensure valid and accurate data 

Once you have your data, you need to analyze it. This is where things can get weedy and technical, but the analysis stage includes issues such as cleaning your data, creating “crosstabs” to explore interesting subgroups and correlations, and summarizing key ideas and findings.

Learn more about the research process

Ultimate Guide to Publishing Survey-Based Research

Learn how to create, publish and amplify your own original research project with this detailed 44-page guide filled with step-by-step instructions and lots of tips