Speaker 1: When we're conducting a survey, we like to make sure we're asking questions that will collect accurate and useful information. But it isn't always obvious what makes a survey question a good survey question. Here are seven tips that can help in developing good survey questions. One, a good question is one where all respondents interpret the question in a consistent manner. That's not always easy to achieve. The question, have you ever been the victim of a serious crime, might be interpreted differently by different people. Is getting your car stolen a serious crime? Probably. What about your bicycle? What about your skateboard? Make sure your questions are as clear and as specific as possible. Two, a good question is also one where people are willing to answer the question. If a survey asks about very personal things, such as sexual behavior or drug use, you might find lots of people unwilling to answer such questions. If you need to ask personal questions, you need to craft those questions carefully and explain to the respondent why you need this information and assure them that you'll keep their responses confidential or even anonymous. Three, a good question is one that a respondent will answer truthfully. Let's take the following question as an example. Do you own a library card? Seems clear enough, and it doesn't seem too personal. But surprisingly, lots of respondents might say yes, even when they don't have a library card. Why is that? It's something called social desirability. Reading books, being well-read is socially desirable. So someone who never goes to the library might not want to admit that they don't have a library card for fear that they might be judged negatively. It might be better to ask the question this way. Some people get books from a library, others buy them from bookstores or online. Do you currently have a library card or not? That might sound a little awkward, but you're more likely to get a more truthful answer. Four, a good question is one that a respondent can actually answer. If you ask someone how much money they spend on groceries in an average month, you might be surprised how many people would say they weren't sure. Make sure people have the knowledge or information to answer your survey question. Five, avoid double-barreled questions. These are questions that ask for a single answer to a question that really has two different parts. For example, let's say you were asked the following question. Do you believe the United States should spend more money on education and less money locking up non-violent criminals? That's a double-barreled question. Respondents who agree with one part but not the other will have trouble answering this question. You should break that question into two separate questions. Six, avoid biased terms or wording. This is probably the hardest thing to do when writing questions. Should you use the term Obamacare or use the Affordable Care Act? Should you use the term welfare or should you say assistance to the poor? It isn't obvious which is the better terminology, but different terms will likely lead to different answers. Surveys will sometimes give one question to half the sample and a slightly different worded question to another half to see how respondents react to the different wording. Seven, pre-test your questions. This is the most important tip. You need to try out your questions on real people and then ask for feedback. Were they confused by a question? Did they answer in a way that surprised you? Pre-testing will help you find out if you have any problems with your questions. Of course, there is no such thing as a perfect survey question. Instead, our goal is to create questions that elicit meaningful responses. Surveys take time and money. We don't want to waste our time and the respondents' time asking bad questions. Following these seven tips can help you craft clear and useful survey questions.