The success of your survey starts with the kind of questions you ask. Bad survey questions make it difficult for you to gather data objectively while good survey questions allow you to collect insightful data that can be used during a systematic investigation.
Many times, it can be difficult to separate bad survey questions from good survey questions and this poses a huge challenge in research. In this article, we will differentiate bad survey questions from good ones to help you create better surveys in the future.
A good survey question is one that allows you to gather clear, unbiased responses from survey respondents. With a good survey question, you can gain clear insights into the thoughts, experiences, and expectations of your target audience for better decision making.
For instance, in market research, good survey questions will provide you with enough information on the buyer’s journey and how well your brand interacts with them across different touchpoints. This would help you optimize your product to suit your needs.
A Likert scale question is a type of survey question that allows you to measure a respondent’s disposition towards specific assertions in a research context. With a Likert scale question, you can find out the extent to which respondents agree or disagree with different statements in your research.
Likert scale questions are made up of 3,5 or 7-point ranking scales that allow you to indicate the degree to which you agree with the statement in question. All Likert scale questions have a mid-point that caters to neutral responses; that is, respondents who neither agree nor disagree with the assertion.
Likert scale questions are one of the best methods of collecting quantitative data in research. A Likert scale question can be unipolar or bipolar. While a unipolar Likert scale question has one pole and measures one quality, a bipolar Likert scale has two opposites and focuses on two qualities.
It is important to understand what makes a Likert scale question one of the best examples of a good survey question. Likert scale questions are really easy to create and they are simple and straight to the point without complexities.
Also, with little or no biases, they help you to capture respondents’ emotions, thoughts, and emotions which translates to valid research data. Likert scale questions also account for variations in people’s thoughts and experiences and represent these adequately.
A dichotomous question is a type of survey question that has only two possible options which are typically parallel; that is, true/false, yes/no, agree/disagree. Survey respondents cannot provide neutral responses to dichotomous questions as there is no allowance for such.
Dichotomous questions are simple, direct, and straight to the point, and they are best used when you need to collect definite responses from your audience. However, because of their direct nature, dichotomous questions can be limiting in nature and prevent respondents from providing enough information in research.
A multiple-choice question is a type of close-ended question that provides respondents with a fixed set of answer-options they can choose from. Depending on the multiple-choice question, respondents are allowed to choose one option or multiple options that they agree with.
Every multiple-choice question is made up of 3 parts which are the stem, answer, and distractors. The stem is the question, the answer is the appropriate response to the question while the distractors are other options listed in the MCQ.
A bad survey question is one that prevents respondents from providing objective answers in research. These questions usually contain several biases that make it difficult for survey respondents to communicate their true thoughts, preferences, and experiences.
Many times, having a poor survey response rate, high survey dropout rate, and subjective research outcomes are indicators of bad survey questions in research. To prevent these, you must know the features of bad survey questions so that you can work on avoiding them.
This is a type of survey question that addresses more than one issue but provides for only one response. It is also known as a double direct question because it weaves multiple issues into one and expects respondents to address these issues with only one answer.
Double-barreled questions are problematic in surveys because they can be confusing and misleading. Weaving multiple assertions into one question that demands a single response makes it nearly impossible for respondents to agree to one assertion and leave out the others.
Examples of Double-Barrelled Questions
Do you find our product interesting and useful?
How satisfied are you with our customer service and service delivery?
A loaded question is a type of question with inherent biases. It typically contains a controversial assumption that typically presupposes that the respondent is guilty of a specific action or behavior. Many times, the assumptions in a loaded question are unverified and this is what makes them problematic.
A loaded question is a bad survey question because it imposes unverified implicit or explicit assumptions on the respondent. This makes it difficult for the respondent to freely communicate his or her thoughts and experiences about the issue at hand.
Examples of Loaded Questions in Research
This question assumes that the respondent smokes which may not be the case. Providing a yes/no answer to this question already confirms the intrinsic bias of the respondent's smoking habit.
b. Do you think we should report this criminal?
This question presupposes that the person in question is a criminal. Whether you answer yes/no to this question, it still confirms the existing bias.
This is a type of survey question that subtly prompts the respondent to provide answers in line with predetermined responses. With a leading question, the researcher already knows what he or she wants you to say so they craft the question to make you respond exactly how they want.
A leading question is a bad survey question because it leads to survey response bias and typically boxes respondents into a corner. This is because it is extremely suggestive. Leading questions can be based on assumptions, interconnected statements, coercion, or direct implications.
Examples of Leading Questions in Survey Research
A negative question is a type of survey question that requires a negative answer for a positive response and a positive answer for a negative response. It is a bad survey question because it is quite complex and can confuse respondents.
As we’ve said earlier, the best kinds of survey questions are simple, straight to the point, and easy to understand. Having a negative question or double negative question in your survey defeats this purpose and can affect the quality of information you gather from your survey.
Examples of Negative Questions in Surveys
A vague question is one that is uncertain or unclear. This type of question does not seek a specific response and it is usually too broad or poorly-defined. Because vague questions are not restricted to a specific context, they always result in generic responses that are not useful in research.
Vague questions almost defeat the purpose of your survey because they make it difficult for you to gather valid data. To avoid vague questions in your survey, ensure that your questions are crafted in a way that they are easy to understand, specific, and well-defined for your audience.
Examples of Vague Questions in Surveys
With Formplus, you can say goodbye to bad online surveys. Our form builder is easy to use and has more than 30 form fields that help you collect and process different types of information without stress. Follow these easy steps to create good online surveys using Formplus.
To get the most out of your survey, of any kind, you must start with the right type of questions. It doesn’t matter whether you have the best survey design or the most effective sharing platforms; bad survey questions will ruin your data collection process.
A lot of effort goes into surveys and data collection and this is why you must strive to get them right at the first trial. Knowing the difference between bad survey questions and good survey questions takes you even closer to having an effective data collection process.
Conduct online surveys with FormplusStart for Free
You may also like:
Collecting data samples in survey research isn’t always colored in black and white. Sometimes, members of your research population may be ...
Developed by Charles Osgood in 1957, the semantic differential scale plays an integral role in helping researchers understand the emotive ...
How do you decide how much a job is worth? 2 ways–you can simply pull a figure off the top of your head and hope that it works or, you can ...
The response rate on your survey is an important indicator of its success. It's not enough to put together the right questions and send out ...