Questionnaires (Leaving Cert Mathematics): Revision Notes
Questionnaires
What are questionnaires?
A questionnaire is a structured set of questions that researchers design to collect specific data from individuals. This method allows researchers to gather information systematically and efficiently from multiple people about their opinions, behaviours, or characteristics.
The individuals who complete questionnaires are called respondents. These are the people whose answers provide the data that researchers need for their studies.
Questionnaires are one of the most widely used data collection methods in research because they can reach large numbers of people quickly and cost-effectively, making them ideal for gathering standardised information.
How questionnaires can be administered
There are two main ways to conduct questionnaires, each with different advantages:
Interview method: An interviewer reads the questions aloud to the respondent and records their answers on the questionnaire form. This approach allows for clarification of questions and ensures all questions are answered.
Self-completion method: Respondents receive the questionnaire and fill in their own responses without an interviewer present. This method can be more cost-effective and may encourage more honest responses to sensitive topics.
Choosing the Right Method
The choice between interview and self-completion methods depends on factors such as budget, time constraints, the complexity of questions, and the sensitivity of the topic being researched.
Guidelines for writing effective questions
When creating questions for a questionnaire, researchers must follow several important principles to ensure they collect reliable and useful data:
Essential Guidelines for Question Writing
-
Be clear about your objectives: Before writing any questions, determine exactly what information you need and what you want to discover from your research.
-
Keep questions short and simple: Complex or lengthy questions can confuse respondents and lead to inaccurate answers. Simple language ensures everyone can understand what is being asked.
-
Start with easy questions: Beginning with straightforward, non-threatening questions helps put respondents at ease and encourages them to continue with the questionnaire.
-
Provide response boxes: Wherever possible, offer tick-boxes or multiple choice options rather than requiring written responses. This makes the questionnaire quicker to complete and easier to analyse.
-
Avoid leading questions: Questions should not suggest what the "correct" answer should be or influence the respondent's choice.
-
Avoid personal questions: Sensitive topics involving names, exact ages, weights, or other personal details should be avoided as they may make respondents uncomfortable or reluctant to participate.
Types of response formats
Multiple-response questions
Multiple-response questions offer respondents a choice of predetermined answers. This format is particularly useful for questions like "What age are you?" where responses can be grouped into categories.
Worked Example: Age Range Categories
A well-designed age question should ensure categories do not overlap:
- ✓ Correct: "Under 18 years", "18-30", "31-50", "Over 50"
- ✗ Incorrect: "Under 18", "18-30", "30-50", "50+" (overlaps at 18, 30, and 50)
Each respondent should find exactly one appropriate category for their situation.
Opinion scales
Opinion scales measure people's attitudes or feelings about particular topics. However, these scales present a common challenge: respondents often cluster their answers around the middle options because they prefer not to appear extreme in their views.
A typical opinion scale might range from "strongly disagree" through "disagree", "no opinion", "agree" to "strongly agree". To combat middle clustering, researchers sometimes use an even number of response options, removing the middle choice and forcing respondents to lean slightly towards agreement or disagreement.
Avoiding bias in questionnaires
Bias occurs when questions are written in ways that might influence responses or make the data unrepresentative. Researchers must carefully avoid several types of problematic questions:
Critical Bias Types to Avoid
Vague questions: Terms like "sometimes", "occasionally", and "often" mean different things to different people. A question asking "How often do you play tennis?" with options "Sometimes", "Occasionally", "Often" is too vague because these words have no clear definitions.
Leading questions: These questions suggest that one particular answer is more acceptable or correct than others. For example, "Normal people enjoy swimming. Do you enjoy swimming?" leads respondents towards answering "yes" because it implies that disagreeing would make them "abnormal".
Personal questions: Questions that ask for sensitive information like exact income, weight, or details about illegal activities often result in dishonest answers or non-response. People may refuse to answer or provide false information to protect their privacy.
Questions with built-in assumptions: Avoid questions like "Do you agree that the EU is now big enough?" because this assumes the EU is already big and suggests agreement is the expected response.
Pilot surveys
Before conducting a full-scale questionnaire study, researchers should always carry out a pilot survey. This is a small-scale trial version of the questionnaire given to a limited number of people.
Purpose of Pilot Surveys
The pilot survey serves several important purposes:
- Identifies problems with question wording that might confuse respondents
- Reveals whether the response options cover all possible answers
- Ensures the questionnaire design will produce the type of information needed for the research objectives
- Allows corrections to be made before the main survey, saving time and resources while improving data quality
Problems discovered during pilot testing can be corrected before the main survey, saving time and resources while improving the quality of data collected.
Key Points to Remember:
-
A questionnaire is a structured set of questions designed to obtain data from individuals, with respondents being the people who answer them
-
Questions can be administered through interviews or self-completion, each method having different advantages depending on the research situation
-
Effective questions must be clear, short, simple, and should avoid being vague, leading, or overly personal
-
Multiple-response questions and opinion scales are useful formats, but opinion scales often suffer from middle clustering bias
-
Always conduct a pilot survey first to identify and fix problems before the main data collection begins