Survey Editing Best Practices

While ideally surveys will be perfect from the moment they are created and long before they have received their first response, surveys are made by human beings, and human beings have a tendency to make mistakes. You may have already completed data collection before you find any number of errors including, but not limited to: Spelling or grammatical errors. Questions that are leading the answer or difficult to understand.

Mistakes in logic branching behind the scenes in the survey. You need to fix these errors when you run the survey in the future, even though changing anything in a survey has the potential to alter the results.

In addition, after the edits are made, there are several things you need to be careful of doing: Assuming New Data is Accurate The biggest problem that researchers make is assuming that the new data is accurate data, and that the edit fixed the problem. Confusion can still exist, questions can be unclear – there are other reasons that your data may still be biased post-edit. Assuming that the data is valid once the change has been made can lead you to bad conclusions.

Edits Can Change Other Data When you edit a single question in your survey, it is possible that your edits change the results of other questions in your survey. How and why is not always known, but it has been found that each question may in some way relate itself to other questions, and changing the question could affect the other questions in the survey. It is also for this reason that you should edit the entirety of the survey all at once, and introduce changes in one big group – making a small change every time you spot an error can cause every survey to get different results.

Edits May Not Be Necessary Depending on the question and the error, an edit may have no purpose. A question that may seem confusing may not have been confusing, or an answer that didn’t seem necessary in retrospect doesn’t harm the survey when kept. Every change can affect everything you collect in the future, and so while editing is important – especially when the error is blatant or occurs before the survey has been run the first time – not all edits are worth completing.

There may be times when you need to edit a survey, but it’s important to remember that an edit in research is not like an edit on a Word document. Small changes can have big effects, and you need to be careful about how and what you change in a survey.