The problem of respondent honesty is a real issue in surveys. All around the world surveys are being taken place for many different reasons but it is so easy to lie on surveys or say you are the person conducting it, it is so easy to receive untrue answers. Now what can you do to get better real answers? Let's say you are conducting a survey on an uncomfortable topic such as abortion. Now everybody is entitled to their own opinion, right? But if you want to receive as much real answers possible, your opinion in this case should not be stated whereas it can lead people from one opinion to the other. It is not a persuasion survey, it should be an unbiased survey where you are showing your audience that your opinion does not matter as well as not shown. The only opinions that should matter is theres. Another thing that can help, is to ask more straight forward questions. Instead of making the question longer and more complicated, cut the questions down to only ask the importance of the question and what really matters most. This can help the audience make a more straight forward response, especially with multiple choice questions. It'll make things less confusing between answer options. You could also run your survey through a couple of people that you know, just so they can look out for any other possible errors you might have missed out on. Rephrasing yes and no questions/responses will help a lot and make things clearer for your audience too. Something that is huge and I have not seen available in surveys very often is having your survey be available in multiple different languages. This can have a huge impact on your audiences answers. It can help increase accurate responses whereas everybody reads, writes, and learns better in different languages. Having your survey available in multiple optional languages can help increase the audiences responses more precisely. 

Correlation is the apparent relationship between two factors. For example, as more gasoline is consumed, more car accidents occur. Causality is being able to explain why such a relationship exists. In this example, causality would be saying that the reason for this is because more gas means there are more cars driving on the road, and more cars driving on the road means there is a greater likelihood of an accident. 

One example of a spurious relationship is the positive relationship between height and reading ability. As a person gets taller, their ability to read and understand language gets better. In reality, height and reading ability are not linked. What does link height and weight, however, is age. As a person gets older, they tend to get taller and at the same time they are progressively learning more about reading comprehension through education.