Surveys are a way of gathering information from people. They can be conducted in many ways such as through a printed questionnaire, over the telephone, in person, or online.
One thing is for sure - we don't want our survey to be biased! Bias is a way of influencing an event in an unfair way. Whether it happens by mistake or on purpose, it means you've cheated a little bit to get your results. Let's look at how to avoid biased results.
Who Should you Ask?
One of the first things to consider when you're creating a survey is who you actually want to answer the survey. Sometimes you may just want a random sample so anyone can answer. However, sometimes, you may want to be more specific. For example, if you wanted students' opinions on changing the school uniform, would it make sense to ask teachers and parents for their opinions too? No because you were only interested in students' opinions.
Discussion
Who would you survey to answer the following questions?
1. What is the most popular type of fruit in the school?
2. Where would students at your school prefer to go on an excursion?
3. Which news program do parents like most?
4. Which sport is the most popular in Australia?
Avoid Dodgy Questions
It's really important to avoid asking leading and ambiguous (unclear) questions when you're creating a survey. Otherwise your results may be misleading, untrue or biased.
Leading questions subtly hint to people that there is a correct way to answer the question. For example, say I asked, "Do you prefer when the weather is nice and warm, or cool?" This question is leading because the word "nice" makes it sound like warm weather is much better than cool weather and, even though people might not agree, they may feel like they should answer that way. A better question to ask would be, "Do you prefer when the weather is warm or cool?" or "What type of weather do you enjoy the most?"
Ambiguous questions are just confusing! For example, say I wanted people to answer yes or no to the question, "Do you like dogs and cats?" How do people who like dogs but not cats answer this question? It would be better if I'd asked two separate questions- "Do you like dogs?" and "Do you like cats?"
Activity
Look at the following leading questions. Discuss why they are leading and how they could be rephrased to remove any bias from a survey.
1. Do you agree that dogs make the best pets?
2. How much do you think the price of gas will go up next year?
3. Do you like soccer and ballet?
4. Do you prefer the packaging with the delicious gourmet meal on the front or the plain packaging?
What Types of Questions can we Ask?
We need to consider the types of questions we want to ask people so we get meaningful data (i.e. data that will help us answer the question we are asking). There are two main types of questions: open-ended and closed-ended questions.
A closed-ended (or closed) question is a question that has a limited number of responses, such as a Yes/No question or a multiple choice question. For example, if I wanted to find out about people's favourite fruits, I may ask, "Which fruit do you prefer- apples, bananas or watermelon?" There are only 3 possible answers people can choose- apples, bananas or watermelon.
An open-ended (or open) question is a question that does not limit people's answers. Using the same example of finding out about people's favourite fruit, I may simply ask, "What is your favourite fruit?" People can answer as they wish.
Most researchers who use surveys use a mixture of open and closed questions because of the pros and cons of each type of question.
| Pros | Cons | |
|---|---|---|
| Open |
Allow people to give detailed answers & explanations. Do not restrict people's answers. May give you an answer that you did not expect. |
May take more time and effort for people to answer the questions. Hard to compare different people's responses. If you end up with too many different responses, it's hard to make judgements/ conclusions from data. |
| Closed |
Quick & easy for people to answer. Easy to compare different people's responses. Easy to make judgements/ conclusions from data. |
People's true choice/ opinion may not be an option, which may give misleading results. People can only give simple responses (no room for explanation). People may not spend as much time or effort answering your questions. |
Questions
1. Can you think of any other pros and cons for using open and closed questions? Write a list.
2. Do you think it's better to use open or closed questions, or a mixture of both? Why?
Time to Design!
Group Activity
1. In groups of 3, design a survey to conduct among your classmates (e.g. favourite colour, favourite sport, favourite book). Think of the type of questions you'd like to ask and write them down.
2. Keep a tally of the results and record your findings.
3. Discuss your findings with your class.