What are some mistakes to avoid in market validation?

If you focus too much on straight survey data, you can fall into a gap.  While surveys can provide valuable insights, they're just one tool. Combine them with in-depth interviews, focus groups, and prototype testing to get a holistic view - also learn to incorporate your data into one set to be able to do larger-scale predictive and prescriptive analytic views.

One mistake to avoid is asking leading questions to customers and people, such as whether they like your product idea, whether it would be useful to them, or other leading questions where they could easily agree or say yes/no. 

A more useful approach is to ask open ended questions that allow them to talk about their needs and problems, and then you can see if your product idea helps address that. Open ended questions about their needs could show you that while yes, your product could be "useful" to them, it's not on their top 5 list of priorities, so they would likely buy other products instead with their limited budget, even if they say they like yours.