How to use survey design (with steps and best practices)
By Indeed Editorial Team
Published 6 April 2022
The Indeed Editorial Team comprises a diverse and talented team of writers, researchers and subject matter experts equipped with Indeed's data and insights to deliver useful tips to help guide your career journey.
Conducting surveys is a standardised way to collect important information in a variety of fields. Surveys are essential to researchers who want to know more about the characteristics, beliefs and preferences of a group of people. Knowing how to conduct effective surveys can help you gather crucial information you require for organisational or personal goals. In this article, we show you how to use survey design, define what a survey is and share best practices in conducting surveys.
How to use survey design
Follow these steps to understand how to use survey design:
1. Define the problem and population
Before you begin a survey, you require a definite research problem. A research problem sets clearly the issue(s) you like to address. Depending on the problem, you can decide who you might invite to take part in the study. The target population refers to the people you're researching. The goal of your survey is to generate results from that microcosm to apply to other people like them.
Since it's practically impossible to study everyone in the target population, you survey a representative sample of the entire population. The size of your sample reflects the size of the population. There are various sampling methods available that enable you to generalise your findings to large populations. Your inferences are likely to be more valid if your sample is larger and more representative.
2. Determine the survey type
The next step is to determine the survey approach to use. Two survey data collection methods you can use are the questionnaire and interview method. A questionnaire is a survey in which you distribute questions to respondents for them to fill it out on their own. You can send the questionnaire by traditional mail, by email or you can drop the questionnaire in person. If your research focuses on a specific location, you can have respondents fill the tool on the spot. For example, you can ask students to fill it at the end of a class.
An interview entails a researcher asking a series of questions via the telephone or in a face-to-face setting as you record the responses. Interviews, like questionnaires, are helpful in collecting quantitative data, where a researcher rates each response and statistically evaluates the results. Interviews are more typically essential in gathering qualitative data, where you transcribe and analyse comprehensive responses of the interview participants to garner a more comprehensive understanding of people's beliefs and opinions.
3. Consider the questions to ask
In this step, you decide the questions to ask and determine the best way to ask them. Factors to consider in your questions are the style of questions to ask, their content, phrasing and order. Survey questions can have open-ended or closed-ended. Closed-ended questions present the respondent with a predetermined set of answers from which to choose. A closed-ended question can entail a binary choice where respondents pick between two choices, such as yes or no. They can also involve a list of options or a scale where respondents select the option that best represents their opinion.
For quantitative research, closed-ended questions work best. They supply numerical data that can you can statistically analyse to discover patterns, correlations and causal links. With open-ended questions, there are no predetermined answers. Rather, the participant expresses their thoughts using their chosen words. Interviews typically use open-ended questions. They can also be helpful in questionnaires to probe further and request respondents to offer a more detailed account of answers to closed questions.
4. Gather research data
In this step, make a detailed research plan. Determine the number of respondents and responses you require and how to arrive at a representative sample. Once you're confident that you've developed a robust research strategy capable of addressing the research problem, you can administer the survey via your preferred method.
5. Analyse data
There are multiple techniques you can use to analyse the survey data. To begin, you process the data, typically using a computer application. You also clean the data to extract any responses that are incomplete. If your research used open-ended questions, it's essential that you start by coding the answers by labelling and organising them into themes. A software application typically assists you in statistical analysis, where you can use the survey data to multiple analyses.
6. Report research findings
Once you've gathered and analysed all the information, you write a report containing your method and findings. In the report, describe how you carried the survey, the sort of questions you asked, sampling methods, time and place of the survey and how many respondents provided information. The questionnaire or interview questions can form part of the document's appendix. Describe the data preparation and statistical methods you used to analyse the data. Summarise findings from your analysis in the results section. Explain and interpret the findings and outline the study's limitations in the discussion section.
What is a survey?
A survey is a set of questions that a researcher or company poses to a group of people to collect their responses and draw inferences relying on those responses. Researchers may use specific survey designs to reach a large audience or a specific demographic. Undertaking a survey is a skill that has a wide range of applications, with professionals from a variety of disciplines using surveys to inform their decisions and find solutions for improving performance. Survey research commonly applies in the following fields:
Social research: Social research is the study of the experiences and characteristics of various social groups.
Market research: Market research entails discovering what consumers think about product offerings, services and brands.
Health research: Health research entails gathering information from patients about their symptoms and treatments.
Politics: Politics is the study of public opinion about political parties and policies.
Psychology: Psychology is the study of personality traits, preferences and behaviours.
Surveys can also be helpful in cross-sectional research, where you collect data only once, and longitudinal research, where you survey the same sample several times over a longer duration.
Best practices in survey design
Here are some best practices to ensure you succeed in a survey:
Compose pertinent questions
The most crucial aspect of the survey process is developing questions that calculate participants' attitudes and opinions. If the information gathered reflects prejudiced questions, the response rate is likely to be low. Composing pertinent questions and using a good research design to organise those questions is a prerequisite for gathering good feedback.
Keep the survey brief
Keep the survey brief and to the point. If a topic isn't precise, it might irritate respondents. Calm and engaged respondents are more likely to complete the survey. Arrange the questions in a logical order and adhere to a particular topic. If the survey is too long, survey participants may lack the motivation to complete the entire study, so survey results may be unreliable.
Ask one question at a time
Branching questions or asking several questions at a time may confuse respondents. They may freeze and select the incorrect option/options from the list of answer options provided because of anxiety. It's best to ask one question at a time to provide clarity to respondents.
Avoid difficult or technical terms
When creating a survey, language is extremely important. The goal is to make responding to your survey as simple as possible for the respondent. The more complicated it is, the more likely the respondents are to abandon their surveys. Avoid using difficult and technical terms to get respondents to respond to the survey. Respondents are likely to understand and respond to surveys better if you write in a language they understand.
Make every question count
Because you're designing your research instrument to elicit critical insights, each question has a direct role in achieving that goal. Ensure that every question provides value and generates survey responses that are directly related to your research objectives. For example, if your participant's age or gender is significant to your results, you can ask this information. If not, you can save your respondents time by skipping it. You can also use multiple-choice questions to elicit responses that are more detailed than a simple yes or no.
Respondents can easily answer questions about their gender and age. Attitudes and opinions on a specific issue may be more of a challenge for some people because they're subjective. Overall, phrase attitudes and opinion questions in a way that best reflects how respondents think and talk about a specific issue so that you can gather data the respondent finds difficult to articulate.
Explore more articles
- What are executive functions? (Plus issues and treatment)
- 7 free email accounts for business and when to use them
- What is firmographic data? (With importance and examples)
- How to calculate the terminal value formula (two methods)
- Data science for managers (11 tips on how to improve)
- Data science vs data analytics: what's the difference?
- Project management vs. programme management: what's the difference?
- How to succeed at marketing business services (with tips)
- 10 important high performing team characteristics to know
- What is coaching in organisations? (With types and benefits)
- Legal and ethical standards: what's the difference?
- Studying marketing in the UK (benefits and modules)