For decades, researchers and businesses have used survey research to produce statistical data and explore ideas. The survey process is simple, ask questions and analyze the responses to make decisions. Data is what makes the difference between a valid and invalid statement and as the American statistician, W. Edwards Deming said:
“Without data, you’re just another person with an opinion.” - W. Edwards Deming
In this article, we will discuss what survey research is, its brief history, types, common uses, benefits, and the step-by-step process of designing a survey.
What is Survey Research
A survey is a research method that is used to collect data from a group of respondents in order to gain insights and information regarding a particular subject. It’s an excellent method to gather opinions and understand how and why people feel a certain way about different situations and contexts.
Brief History of Survey Research
Survey research may have its roots in the American and English “social surveys” conducted around the turn of the 20th century. The surveys were mainly conducted by researchers and reformers to document the extent of social issues such as poverty. (1) Despite being a relatively young field to many scientific domains, survey research has experienced three stages of development (2):
- First Era (1930-1960)
- Second Era (1960-1990)
- Third Era (1990 onwards)
Over the years, survey research adapted to the changing times and technologies. By exploiting the latest technologies, researchers can gain access to the right population from anywhere in the world, analyze the data like never before, and extract useful information.
Survey Research Methods & Types
Survey research can be classified into seven categories based on objective, data sources, methodology, deployment method, and frequency of deployment.
Surveys based on Objective
Exploratory Survey Research
Exploratory survey research is aimed at diving deeper into research subjects and finding out more about their context. It’s important for marketing or business strategy and the focus is to discover ideas and insights instead of gathering statistical data.
Generally, exploratory survey research is composed of open-ended questions that allow respondents to express their thoughts and perspectives. The final responses present information from various sources that can lead to fresh initiatives.
Predictive Survey Research
Predictive survey research is also called causal survey research. It’s preplanned, structured, and quantitative in nature. It’s often referred to as conclusive research as it tries to explain the cause-and-effect relationship between different variables. The objective is to understand which variables are causes and which are effects and the nature of the relationship between both variables.
Descriptive Survey Research
Descriptive survey research is largely observational and is ideal for gathering numeric data. Due to its quantitative nature, it’s often compared to exploratory survey research. The difference between the two is that descriptive research is structured and pre-planned.
The idea behind descriptive research is to describe the mindset and opinion of a particular group of people on a given subject. The questions are every day multiple choices and users must choose from predefined categories. With predefined choices, you don’t get unique insights, rather, statistically inferable data.
Survey Research Types based on Concept Testing
Monadic Concept Testing
Monadic testing is a survey research methodology in which the respondents are split into multiple groups and ask each group questions about a separate concept in isolation. Generally, monadic surveys are hyper-focused on a particular concept and shorter in duration. The important thing in monadic surveys is to avoid getting off-topic or exhausting the respondents with too many questions.
Sequential Monadic Concept Testing
Another approach to monadic testing is sequential monadic testing. In sequential monadic surveys, groups of respondents are surveyed in isolation. However, instead of surveying three groups on three different concepts, the researchers survey the same groups of people on three distinct concepts one after another. In a sequential monadic survey, at least two topics are included (in random order), and the same questions are asked for each concept to eliminate bias.
Based on Data Source
Primary Data
Data obtained directly from the source or target population is referred to as primary survey data. When it comes to primary data collection, researchers usually devise a set of questions and invite people with knowledge of the subject to respond. The main sources of primary data are interviews, questionnaires, surveys, and observation methods.
Compared to secondary data, primary data is gathered from first-hand sources and is more reliable. However, the process of primary data collection is both costly and time-consuming.
Secondary Data
Survey research is generally used to collect first-hand information from a respondent. However, surveys can also be designed to collect and process secondary data. It’s collected from third-party sources or primary sources in the past.
This type of data is usually generic, readily available, and cheaper than primary data collection. Some common sources of secondary data are books, data collected from older surveys, online data, and data from government archives. Beware that you might compromise the validity of your findings if you end up with irrelevant or inflated data.
Based on Research Method
Quantitative Research
Quantitative research is a popular research methodology that is used to collect numeric data in a systematic investigation. It’s frequently used in research contexts where statistical data is required, such as sciences or social sciences. Quantitative research methods include polls, systematic observations, and face-to-face interviews.
Qualitative Research
Qualitative research is a research methodology where you collect non-numeric data from research participants. In this context, the participants are not restricted to a specific system and provide open-ended information. Some common qualitative research methods include focus groups, one-on-one interviews, observations, and case studies.
Based on Deployment Method
Online Surveys
With technology advancing rapidly, the most popular method of survey research is an online survey. With the internet, you can not only reach a broader audience but also design and customize a survey and deploy it from anywhere. Online surveys have outperformed offline survey methods as they are less expensive and allow researchers to easily collect and analyze data from a large sample.
Paper or Print Surveys
As the name suggests, paper or print surveys use the traditional paper and pencil approach to collect data. Before the invention of computers, paper surveys were the survey method of choice.
Though many would assume that surveys are no longer conducted on paper, it's still a reliable method of collecting information during field research and data collection. However, unlike online surveys, paper surveys are expensive and require extra human resources.
Telephonic Surveys
Telephonic surveys are conducted over telephones where a researcher asks a series of questions to the respondent on the other end. Contacting respondents over a telephone requires less effort, human resources, and is less expensive.
What makes telephonic surveys debatable is that people are often reluctant in giving information over a phone call. Additionally, the success of such surveys depends largely on whether people are willing to invest their time on a phone call answering questions.
One-on-one Surveys
One-on-one surveys also known as face-to-face surveys are interviews where the researcher and respondent. Interacting directly with the respondent introduces the human factor into the survey.
Face-to-face interviews are useful when the researcher wants to discuss something personal with the respondent. The response rates in such surveys are always higher as the interview is being conducted in person. However, these surveys are quite expensive and the success of these depends on the knowledge and experience of the researcher.
Based on Distribution
The easiest and most common way of conducting online surveys is sending out an email. Sending out surveys via emails has a higher response rate as your target audience already knows about your brand and is likely to engage.
Buy Survey Responses
Purchasing survey responses also yields higher responses as the responders signed up for the survey. Businesses often purchase survey samples to conduct extensive research. Here, the target audience is often pre-screened to check if they're qualified to take part in the research.
Embedding Survey on a Website
Embedding surveys on a website is another excellent way to collect information. It allows your website visitors to take part in a survey without ever leaving the website and can be done while a person is entering or exiting the website.
Post the Survey on Social Media
Social media is an excellent medium to reach abroad range of audiences. You can publish your survey as a link on social media and people who are following the brand can take part and answer questions.
Based on Frequency of Deployment
Cross-sectional Studies
Cross-sectional studies are administered to a small sample from a large population within a short period of time. This provides researchers a peek into what the respondents are thinking at a given time. The surveys are usually short, precise, and specific to a particular situation.
Longitudinal Surveys
Longitudinal surveys are an extension of cross-sectional studies where researchers make an observation and collect data over extended periods of time. This type of survey can be further divided into three types:
- Trend surveys are employed to allow researchers to understand the change in the thought process of the respondents over some time.
- Panel surveys are administered to the same group of people over multiple years. These are usually expensive and researchers must stick to their panel to gather unbiased opinions.
- In cohort surveys, researchers identify a specific category of people and regularly survey them. Unlike panel surveys, the same people do not need to take part over the years, but each individual must fall into the researcher’s primary interest category.
Retrospective Survey
Retrospective surveys allow researchers to ask questions to gather data about past events and beliefs of the respondents. Since retrospective surveys also require years of data, they are similar to the longitudinal survey, except retrospective surveys are shorter and less expensive.
Why Should You Conduct Research Surveys?
“In God we trust. All others must bring data” - W. Edwards Deming
In the information age, survey research is of utmost importance and essential for understanding the opinion of your target population. Whether you’re launching a new product or conducting a social survey, the tool can be used to collect specific information from a defined set of respondents. The data collected via surveys can be further used by organizations to make informed decisions.
Furthermore, compared to other research methods, surveys are relatively inexpensive even if you’re giving out incentives. Compared to the older methods such as telephonic or paper surveys, online surveys have a smaller cost and the number of responses is higher.
What makes surveys useful is that they describe the characteristics of a large population. With a larger sample size, you can rely on getting more accurate results. However, you also need honest and open answers for accurate results. Since surveys are also anonymous and the responses remain confidential, respondents provide candid and accurate answers.
Common Uses of a Survey
Surveys are widely used in many sectors, but the most common uses of the survey research include:
- Market research: surveying a potential market to understand customer needs, preferences, and market demand.
- Customer Satisfaction: finding out your customer’s opinions about your services, products, or companies.
- Social research: investigating the characteristics and experiences of various social groups.
- Health research: collecting data about patients’ symptoms and treatments.
- Politics: evaluating public opinion regarding policies and political parties.
- Psychology: exploring personality traits, behaviors, and preferences.
6 Steps to Conduct Survey Research
An organization, person, or company conducts a survey when they need the information to make a decision but have insufficient data on hand. Following are six simple steps that can help you design a great survey.
Step 1: Objective of the Survey
The first step in survey research is defining an objective. The objective helps you define your target population and samples. The target population is the specific group of people you want to collect data from and since it’s rarely possible to survey the entire population, we target a specific sample from it. Defining a survey objective also benefits your respondents by helping them understand the reason behind the survey.
Step 2: Number of Questions
The number of questions or the size of the survey depends on the survey objective. However, it’s important to ensure that there are no redundant queries and the questions are in a logical order. Rephrased and repeated questions in a survey are almost as frustrating as in real life. For a higher completion rate, keep the questionnaire small so that the respondents stay engaged to the very end. The ideal length of an interview is less than 15 minutes. (2)
Step 3: Language and Voice of Questions
While designing a survey, you may feel compelled to use fancy language. However, remember that difficult language is associated with higher survey dropout rates. You need to speak to the respondent in a clear, concise, and neutral manner, and ask simple questions. If your survey respondents are bilingual, then adding an option to translate your questions into another language can also prove beneficial.
Step 4: Type of Questions
In a survey, you can include any type of questions and even both closed-ended or open-ended questions. However, opt for the question types that are the easiest to understand for the respondents, and offer the most value. For example, compared to open-ended questions, people prefer to answer close-ended questions such as MCQs (multiple choice questions)and NPS (net promoter score) questions.
Step 5: User Experience
Designing a great survey is about more than just questions. A lot of researchers underestimate the importance of user experience and how it affects their response and completion rates. An inconsistent, difficult-to-navigate survey with technical errors and poor color choice is unappealing for the respondents. Make sure that your survey is easy to navigate for everyone and if you’re using rating scales, they remain consistent throughout the research study.
Additionally, don’t forget to design a good survey experience for both mobile and desktop users. According to Pew Research Center, nearly half of the smartphone users access the internet mainly from their mobile phones and 14 percent of American adults are smartphone-only internet users. (3)
Step 6: Survey Logic
Last but not least, logic is another critical aspect of the survey design. If the survey logic is flawed, respondents may not continue in the right direction. Make sure to test the logic to ensure that selecting one answer leads to the next logical question instead of a series of unrelated queries.
How to Effectively Use Survey Research with Starlight Analytics
Designing and conducting a survey is almost as much science as it is an art. To craft great survey research, you need technical skills, consider the psychological elements, and have a broad understanding of marketing.
The ultimate goal of the survey is to ask the right questions in the right manner to acquire the right results.
Bringing a new product to the market is a long process and requires a lot of research and analysis. In your journey to gather information or ideas for your business, Starlight Analytics can be an excellent guide. Starlight Analytics' product concept testing helps you measure your product's market demand and refine product features and benefits so you can launch with confidence. The process starts with custom research to design the survey according to your needs, execute the survey, and deliver the key insights on time.
References
- Survey research in the United States: roots and emergence, 1890-1960 https://searchworks.stanford.edu/view/10733873
- How to create a survey questionnaire that gets great responses https://luc.id/knowledgehub/how-to-create-a-survey-questionnaire-that-gets-great-responses/
- Internet/broadband fact sheet https://www.pewresearch.org/internet/fact-sheet/internet-broadband/