What Is Survey Research?

When you ask thematic questions to a group of people and then analyze the results, you conduct survey research. You need to remember that successful research of this type depends on your ability to define the participants of a survey and the type of it (in person, by mail, or online), compose the layout and questions, and distribute the questions of the survey among your target group, analyze the answers, and note the results.

Surveys are an efficient method of data collection. They are flexible and can be used in various types of research designs.

Why Do You Need Surveys?

You can use surveys for research in different fields of study to learn more about the most prominent characteristics, opinions, beliefs or misbeliefs, and preferences of the target groups. For example, you can use surveys in:

• market research - to learn customers’ opinions about products, brands, and services
• politics - to evaluate public attitudes toward governmental policies, movements, and parties
• social research - to find out more information about important characteristics, trends, and experiences of social groups
• psychology - to research behaviors, traits of character, and personality layouts
• health research - to collect data about symptoms, methods of treatment, and their efficiency among the patients of a target group

You can use surveys in different types of studies, such as cross-sectional and longitudinal ones. In cross-sectional studies, you gather the information you need only once. While in longitudinal studies, you offer several surveys to the same group of people at different times.

How to Conduct a Survey?

There are several steps you need to take to conduct a deep and useful survey for your research.

1. Identifying Populations and Samples

Before the start of a survey, you need to know who you are going to address your questions to. You also need to formulate a clear research question to be sure that you organize your survey work correctly. Only after you make up this question may you think of the target groups your survey will be meant for.

Target Population

It is a specific group of people you want to learn some information from. The group can be either broad or narrow. For example, the population of Canada is a broad group (population), while customers of a shoe-repairing workshop aged 45-60 make up a narrow population.

You have to design the survey in a way to receive results from the entire population but not only a part of it. So, think in advance about what kind of conclusions you are expecting to make, and then think about the size of your population.

Research Sample

If a group of people is too broad to conduct a questionnaire with all of them, for example, people of Canada, you won’t be able to get responses from everyone. So, you need just a sample of this population for your research. Its size will depend on the size of the entire population. You may need an online sample calculator to find out how many responses will be necessary for your research.

If you want to generalize a broad population, use one of the sampling methods to define the part of the population that will represent it all. If your sample is large, you have a chance to receive more reliable conclusions.

2. Choosing the Type of Survey

You can use two types of surveys - a questionnaire and an interview. You may choose one according to the size and location of the sample.

Questionnaires

Sending a survey by mail when you need to collect demographic information has its advantages and disadvantages:

• ✔️ Large samples are easily accessible.
• ✔️ You can control the samples yourself.
• ❌ The rate of responses is pretty low.

When you are a student who needs to do dissertation research, you may opt for an online survey. This method is cheap and flexible. You can utilize such instruments as Google Forms or SurveyMonkey for it. The pros and cons of online surveys are the following:

• ✔️ You are not limited by location or time to reach large samples.
• ✔️ The obtained data is easy to process.
• ❌ Online surveys are mostly anonymous, so you cannot fully control the responses.

When you do a survey in person, you can immediately focus on the location and offer written questionnaires to people you want to participate. For instance, you can enter a school and ask students there to answer your questions on the spot.

There are also several benefits and drawbacks of such a method. They are:

• ✔️ You can meet people in person, so you are sure they are exactly those representatives of the sample you need.
• ✔️ You can ask questions at certain times and locations, for example, at school after the classes are over or before them.
• ❌ You may not be able to ask the people you are interested in because they may be short of time or unwilling to answer your questions.

Interviews

Interviews are conducted orally. This method works very well when the size of a sample is small. You can collect more detailed information, including people’ preferences and opinions. Interviews are often held either in person or by phone.

The pros and cons of this technique involve the following:

• ✔️ Personal contact with people from the sample provide a better understanding of their needs and feelings.
• ✔️ If a person does not understand the question, you can explain it to them and ask for additional information.
• ❌ The interview is not anonymous, so a person may not want to answer all the questions honestly.

Interviews can be used for collecting both quantitative and qualitative data.

3. Think of the Appropriate Questions for a Survey

When you start composing the survey questions, consider the following:

• question types
• phrasing
• order and layout of a survey
• content

Questions can be closed-ended and open-ended. Some surveys can include both of these types.

Closed-ended questions provide a ready set of answers, such as ‘Yes/No’ or ‘Agree/Disagree.’ Likert scales also belong to this type of question. You can also make up a list of variants with only one correct answer (for example, ethnicity) or several correct options, as in the case with preferable lunch foods. This type of question can be used for quantitative research because you can analyze the obtained statistical data to distinguish patterns or correlations.

You should use open-ended questions to do qualitative research. Make a narrow focus on the questions to get consistent answers. Such questions work well in interviews, and they can also be follow-up questions to get more details.

Think carefully about every question’s content. Do not make questions unrelated to the purpose of a questionnaire or interview. In closed-ended questions, the variants of answers should involve all possible answers. You can add ‘other’ to ensure that you get relevant results.

Arrange all the questions in a logical order. Start with closed-ended questions that are easy to answer. If you need to include several topics in your survey, group the questions by themes, forming specific sections. If the response to one question is related to the previous one, let these questions follow each other.

4. Offer the Survey to Your Target Group and Get the Answers

Decide beforehand who will participate in your survey and where and when you are going to conduct it. Think about the type of survey in advance and decide how many questions you want to ask and how many answers you need to get from one sample. Then, think about ways of doing your survey research – in person, online, or by mail.

5. Make an Analysis of the Results

To start analyzing results, you have to process data. You can use a special computer program to sort out the responses. Remove all the answers that are incorrect or incomplete.

Assign labels to the answers to open-ended questions to arrange them into categories. Use thematic analysis that is especially useful in processing interview results. You can use Stata or SPSS software.

6. Note Down the Results

The results of any survey can become a part of your research paper or dissertation. You can describe your survey in the methodology section, including information about the types of questions, methods you have used, and the location and time of the survey. You should also include the response rates and a complete questionnaire in the Appendix.

Then, you have to describe how you processed the data and what statistical methods were used. Make a summary of all the results in the result section and explain them in the discussion and conclusion. Here, you can already answer the research question and provide your comments and interpretations. Highlight the limitations of the research, too.

How to Design a Questionnaire?

All questionnaires consist of a set of specific questions about the target respondents’ opinions and experiences. Questionnaire results are used in both qualitative and quantitative analysis. They are normally used in social and health sciences or in marketing.

In fact, a questionnaire is a potent instrument for gathering data. When you create valid questions connected with your research objective, you have to place them in the correct order and select a proper method for response analysis. Like in surveys, you need to develop a design, define the population, choose the data processing methods, and interpret the information correctly.

Opt for the method of administration - either researcher-administered or self-administered.

Self-administered questionnaires are meant for obtaining deeper and more detailed insights, while self-administered ones are very easy and cheap. You can deliver self-administered questionnaires online, in person, or by email. All the participants should receive the same questions. There are specific advantages and disadvantages of these questionnaires because they can be:

• ✔️ applied to both small and large groups
• ✔️ controlled by their pace
• ✔️ cost-effective
• ✔️ anonymous and applicable to sensitive information
• ❌ difficult for people with limited skills
• ❌ creating a bias towards volunteers who make survey requests
• ❌ usually ignored

Researcher-administered questionnaires are normally applied in the form of interviews held in person, online, or by phone. Let’s consider the pros and cons of such questionnaires since they:

• ✔️ can clarify unclear questions and answer and make them easier to understand and classify them
• ✔️ get more responses because personal communication attracts most people
• ✔️ help consider only those participants that belong to your target group
• ❌ create difficulties for analysis because they are mostly qualitative
• ❌ create bias because of the lack of anonymity
• ❌ take a lot of time and effort
• ❌ can produce bias on the part of a researcher

Questionnaires use the same open-ended and closed-ended questions as other types of surveys do.

A Step-by-Step Guide

To create a consistent questionnaire, you can follow these steps:

1. Identify All the Goals and Objectives

The aims of the questionnaire can be answers to the following questions:

• What would you like to know?
• What items or experiences do you intend to study?
• Is a self-report questionnaire an appropriate method for your research, and why/why not?

Now, you can start thinking about operationalizing the variables and concepts. It means that you need to convert abstract ideas into specific measurements.

2. Use Suitable Questions

You should think in the way your potential respondents do to create good questions. Consider their language levels and the time they can spend answering your questions. Answer these questions:

• Do respondents know the language and terms used in your questionnaire?
• Do response options include all the possible variants?
• Can any questions from the set embarrass or insult the people?
• Do the responses comply with each other, or are they mutually inacceptable?
• Will the participants have enough time to respond to all the open-ended questions?

Ensure that you have included all the possible answer options. The absence of the variant that suits them can distract respondents from answering further questions.

3. How to Order Questions and How Long Can They Be?

If your respondents do not receive any benefits from answering the questionnaire, keep it as short and easy as possible. In other cases, your sample will include only highly motivated respondents, and the answers may be biased.

The order of questions should correspond to the aims. Participants may be short of time, so keep your questionnaire logical. Randomize the questions to avoid bias. However, it will take you more time to analyze the answers.

4. Do the Pretest

You need to see how your questionnaire works and whether everything is clear. Ask your friends or family to answer the questions. Your classmates or colleagues can do it too. After they complete the answers, ask them what was difficult or seemed ambiguous for them. You can also find out whether the directions were consistent and straightforward enough.

How to Design Likert Scales

Likert scales are often used to collect data about opinions, habits, and attitudes. This method is popular for surveys because it allows collecting data about characteristics and emotions that can be analyzed and used for conclusions.

Such a scale consists of a question or statement and several answers (as a rule, 5-7) to choose from. You can also give each answer a numerical score to be able to analyze the data in quantities. The questions can relate to frequencies, satisfaction, or agreement.

Creating Likert-Type Questions

Each question on the scale should measure one part of the topic. You can use either questions or statements here. They should be easy to understand and bias-free.

How to Pick Out Response Items

The items on each side of the scale are known as response anchors. The midpoint out of 5 or 7 items is always neutral. Give each item a score from 1 to 5 or from 1 to 7.

Though you may think that more items can provide more consistent responses, it is not quite right. When people see too many items for making a choice, they feel confused. On the other hand, too few items will not provide you with the details you need. That is why the odd numbers of 5 or 7 are pretty consistent. For example:

• How frequently do you watch horror films?
  - Never
  - Hardly ever
  - Often
  - Usually
  - Always
• How frequently do you go to the cinema?
  - Never
  - Occasionally
  - Sometimes
  - Often
  - Very Often
  - Usually
  - Always

What Are the Main Types of Items?

Likert scales allow for measuring a wide range of emotions or perceptions. They may involve agreement, quality, experience, or likelihood.

Items can be unipolar and bipolar. Unipolar items can measure only one quality (e.g., frequency), while bipolar items relate to two opposite attributes, such as agreement and disagreement.

Your choice will depend on the research aims and questions. If you need a wider range of details, use the bipolar items. If you want to obtain finer details, make them unipolar.

All the items should exclude each other. Do not overlap the items because it will confuse respondents. For example, never use the responses ‘neither agree nor disagree’ and ‘indifferent’ simultaneously.

Likert scales have their pluses and minuses. They are the following:

• ✔️ They allow for quantitative analysis and help statistical hypothesis testing.
• ✔️ The scales provide detailed information about opinions, characteristics, or behaviors.
• ❌ People may avoid selecting extreme points for answers that can cause response bias.
• ❌ If there are many questions on the scales, respondents may get inattentive and tired.
• ❌ They may interpret the vague items subjectively.
• ❌ The closed-ended nature of these questions may result in participants’ unwillingness to make a choice because they do not see the option that corresponds to their feelings at the moment.

Final Thoughts

Now, you know what surveys are, how to make them out and implement them, and what methods and techniques to use when composing them. Use different types of questions, such as open-ended (for qualitative analysis) and closed-ended (for quantitative analysis). You can also opt for questionnaires and interviews conducted in-person, online, or by email.

You can make the most effective choice according to your research question, hypotheses, and research aims. The information provided in this article will help you, so you can reread it whenever you need.