In Class: Confusing Concepts from Logos

One of the things that confused me in Chapter 4 of Everything's an Argument is the use of surveys and polls as a logical appeal. When I think of surveys and polls I think of a very inaccurate measure of the public's feelings. There are many biases surrounding the use of surveys and I feel that they are not a good representative of a logical appeal. My question is, why did they include this as an example in the chapter if the readings and data obtained from the source are inaccurate? I realize polls are used to persuade people towards a certain view on the topic that is presented but would the inaccuracy deter people from using this form to appeal to people?