When it comes to the importance of design, former IBM President Thomas Watson Jr. cut to the chase, "Good design is good business," he said.
Indeed, how things are designed matters. That goes for the personal computer on your desk, the watch on your wrist—or if your focus is market research, the design of your study or survey.
Effective research design can add real value by creating the framework for successfully executing market research in ways that produce the most relevant results for what you aim to achieve.
Research design is the method you choose to set up a study to allow you to gain a greater understanding of an issue. You start with a theory, problem, or idea, and then develop a hypothesis. Your research design will include a systematic way to collect, measure, and analyze data with the aim of reaching a logical conclusion to your hypothesis. Ideally, the results you end up with should be objective, measurable, and actionable.
Prep for success
Preparation is key when it comes to effective research design. Putting careful thought upfront into design improves your chances of the study running smoothly and delivering the most relevant, easy-to-assess, and statistically sound results. Importantly, it also limits the risk of research flaws, bias, or false assumptions.
The best place to start is by clearly outlining the purpose of the study, research design methods, and who will make up the study participants.
A good research design includes:
The result will be well-designed research to clearly answer the hypothesis. The research study’s findings will be objective and reliable, able to be validated, and can be generalized to situations beyond the study group.
Fortunately, there is a wide range of research design options to choose from to land on the just-right choice for your study. Again, this is where careful assessment at the beginning of the process can pay big dividends.
Developing a clear approach to your research helps guide you in this process. Once you have chosen a topic, problem, or hypothesis for your research, ask a series of questions to narrow down which design options are likely the best fit. Questions such as:
Detailed answers to these questions help narrow down the research design that is best for your study.
Another key factor to consider is your timeline. If you only have a few weeks, then it makes sense to select a design in which you can gather and analyze data quickly. Surveys are one great way to quickly gather data from a large population.
If you are collecting detailed information and insights or are focused on a more narrow target audience, you will need more time to identify respondents, collect data, and analyze the results. Observation, interviews, focus groups, or experiments will provide deeper insights, but take longer to administer.
Once you’ve nailed down your research design purpose and timeline, you’re ready to choose the type of research design that will help you achieve your goals. You can assess your choices from the eight research options described below.
If you flip on a light switch, a light will turn on.
That’s a simple example of cause and effect between two things, and that’s what experimental research design aims to find out.
Experimental design studies the relationship between a variable that can be changed, the “independent variable,” and a variable that remains constant, the “dependent variable.” In experimental research design, independent variables are frequently changed to understand potential different effects on the dependent variable.
A common market research example is using experimental research to understand how a product’s price (independent variable) influences a customer (dependent variable) when purchasing a product.
Words can also influence people’s behavior. For example, take these questions:
“Would you be willing to help by giving a donation?”
“Would you be willing to help by giving a donation? Every penny will help.”
Notice how that last sentence could affect someone’s decision to donate? The independent variable (The wording of the request) was changed to understand the effect on the dependent variable (people). In one study it was noted that by adding the words such as “every penny will help” twice as many donations were collected.
The pros of experimental research design include flexibility in adjusting variables, creating two groups to study (an experimental group and control group), and gathering measurements over a long time period.
The potential cons of experimental research design are that it can be time-consuming and expensive. Further, the research results may not be applicable to the general population if the study assumes everyone falls into one or two categories.
Descriptive research design makes sense when there’s limited information available about the problem you’re exploring. That means you need to gather, analyze, and present data to clarify the issue.
A classic example is renowned researcher Jane Goodall who performed a descriptive research study by observing and living with gorillas in Africa. Through her careful observations, Goodall became one of the first scientists to gather extensive, previously unknown information and insights on animal behavior.
A case study, field or lab observation, and surveys are effective ways to perform descriptive research. The advantage of descriptive research design is that it can help you understand a trend, or discover the “when, what, or where” of a situation. Ultimately, after conducting your descriptive research design survey, you can assess why further research is needed.
On the downside, descriptive research design often lacks historical data if the situation has not been previously studied.
Survey research design is a quick way to gather credible information from groups of individuals. Online surveys in particular can quickly tap into a pool of survey respondents to collect extensive and actionable data on consumer behavior, employee performance, buyer preferences, and other insights.
Surveys offer a flexible, quantitative approach to research design. They are often the go-to approach for many organizations to capture data and insights on customers, or study commercial, political, social, psychological, health, and other issues.
Survey results are statistically meaningful and can be targeted at specific groups that match the research goal and design. Manufacturers, governments, employers, and other organizations use surveys to collect valuable insights from a target audience. They use surveys to test product concepts, understand consumer behavior, or gain insights into employee performance.
The advantage of surveys is that they collect self-reported data directly from the respondent as opposed to data collected by the researcher. Not only are survey respondents easily accessible, but also surveys can reach a large audience at a lower cost, allowing you to generalize the results to a larger population.
The disadvantage of online surveys is that the researcher often cannot develop an ongoing relationship with respondents. The researcher may also need to ask many types of questions or go into greater depth than a short online survey.
Qualitative research design involves studying data that doesn’t have quantifiable metrics attached to it. This approach is typically used to study a smaller number of people to gain an in-depth understanding of their behaviors or attitudes.
Often, qualitative research studies gather data from observation. You may observe an event, groups of people, how products are being used, and summarize what you observe. In the context of surveys, qualitative research is often conducted by asking respondents to answer questions in their own words rather than selecting from a predetermined list of answer options.
While the results can be subjective, they can be extremely useful in guiding future research or unearthing new ideas or questions to pursue. But be careful: A common pitfall of qualitative research is assuming that the results of a small number of participants will apply to a larger group.
Quantitative research focuses on statistical evidence. Numerical data is more easily interpreted than qualitative data and easier to compare over time. Quantitative research results are increasingly being used for managerial decision making because it provides a broader, statistical view of current problems.
Often, the data collected can be large and cumbersome, so experience in managing large amounts of data is essential. However, today’s data analytics provide easier-to-read dashboards and reporting that make interpreting data simpler and less time-consuming. Marketers often use quantitative research to determine market strategy and execution, and evaluate marketing effectiveness.
The advantage of quantitative data is that it provides statistical evidence of a theory, behavior, or trend.
The disadvantage of quantitative data is that it’s sometimes difficult to collect a large enough sample size to accurately reflect the general population. Further, it does not provide the breadth of information that qualitative observations or other types of in-depth research offer.
Correlational research design determines how closely two variables are connected. It offers statistical analysis about the relationship but offers no assumptions or hypotheses about the relationship.
A correlation coefficient measures the relationship, ranging from -1 to +1. The closer to +1, the more closely the variables are related. The closer to -1, the greater the indication of a negative relationship.
One of the most direct correlations in economics is the relationship between price and demand. If demand for a product increases, the price will typically go up and vice versa. In the context of correlational research design, price and demand have a positive correlational relationship.
The advantage of correlational research design is that it provides clear evidence of a positive or negative relationship. The study can demonstrate the need for further research.
The disadvantage of correlational research design is that it can be time-consuming. Further, it offers no insight as to which variable is more important in the relationship.
Diagnostic research design helps you discover the underlying cause of an issue which can then lead to a potential solution. This approach can help in instances in which you are unsure why your customers are behaving in a certain way and need more information in order to address a particular challenge or make the most of an opportunity.
Diagnostic research is conducted in three stages: the inception of a problem, diagnosis of the problem, and the solution of the problem. This approach is often used in the healthcare space for clinical trials to understand the effectiveness of new treatments and therapies for specific diseases. It can also be used to address and overcome challenges users have with a product or service.
The advantage of diagnostic research design is that you can uncover a root cause of an issue that can lead to potential solutions. You can choose which variables are affecting behavior and test for a relationship.
The disadvantage of diagnostic research design is choosing a small sample size or the wrong variables to test.
Explanatory research design, also called causal research design, is used to try to understand the cause and effect of variables in your study. While descriptive research designs can show whether a phenomenon occurred, explanatory research design explains “why” it occurred by looking for quantifiable data that show the phenomenon’s causes or reasons.
Simply put, explanatory research seeks to find explanations for why something occurs.
Once you choose a research design, then it is time to put it into action and choose how you will collect data. Below are four types of studies you can use to support the research design you’ve chosen.
Cohort studies allow you to collect data over time about a group of people who have things in common.
Cohorts are groups of people with similar characteristics. For instance, study subjects may live in the same town, be part of the same age group, and have dined at a restaurant within the last month. So, for instance, people between the ages of 46 to 54 who live in Northern California and recently dined at a restaurant would be part of a cohort study.
Cohort studies are often used in medical and social sciences to study groups of people. These groups are studied across time, both retrospectively (analyzing the past) and prospectively (tracking them into the future).
In a business context, a cohort study can focus on a set of users who all use or consume the same product. In a retrospective study, the researcher may track users from the date of their first purchase. They may perform a prospective study to analyze the potential revenue from a customer acquired in the future. These studies help business decision-makers gain deeper clarity about the users of their products and services.
Looking for international respondents for your market research studies? Learn how we gather responses from all over the world.
Cross-sectional studies collect data from their target audience at a specific point in time. The researchers do not change any variables or get comparison data.
You observe people and collect data from those observations. The purpose is to draw conclusions about the reactions you see as well as variables in the study. A cross-sectional research study is not a random sample, but one designed to get the input from a target audience at one point in time.
The advantage of cross-sectional research is that you can use online surveys, making it quick and inexpensive to administer. Online surveys help you capture data from a large audience and draw valid conclusions.
The disadvantage of cross-sectional research is that it is not used to find an underlying cause or understand how different variables relate to each other. There is no follow-up or comparison to other time periods from the same target audience.
Longitudinal studies collect observations, over time, from the same group of people.
For instance, the same group of people may be interviewed or surveyed at regular intervals. This allows researchers to study the impact of variables over time and explain any changes that surface.
Panel studies are a type of longitudinal study that’s focused on asking shorter research questions. The same audience is surveyed at different points in time to measure changes in behavior, emerging trends, and determine the reason for changes.
Panel studies can reveal customer behavior changes to brand repositioning, price changes, packaging redesign, and other variables. Online surveys are frequently used to collect longitudinal data because they are cost-effective and reach a large, targeted audience.
The advantage of longitudinal studies is that it allows researchers to study a problem or theory over time using the same audience. Researchers can shift variables, observe resulting changes, and make conclusions about future behavior.
The disadvantage of longitudinal studies is that if the study is conducted over an extensive time period (many years), it may be increasingly difficult to track variables and their impact. Current trends may change abruptly, so not all conclusions about respondents are accurate for the long term.
The cross-sequential study combines the approach used by longitudinal studies that take place over time and a cross-sectional study that looks at different people all at once.
Cross-sequential studies focus on groups of different people across time. You should also monitor for cohort effects of trends within a similar group of people.
The advantage of cross-sequential design is that researchers can examine large groups of people in a short period of time.
The disadvantage of cross-sequential design is that they are complex and must be monitored more closely.
You should determine your research design strategy before you start collecting data to ensure you get quality results. Here are four tips to make sure you make the most of your research efforts.
What is the purpose of your study? A clearly stated problem, hypothesis, or theory creates a theme for your research study. This can be a problem you believe needs to be solved, an idea that requires more research, or a hypothesis of how consumers will react to a certain product release or policy, or service change.
For instance, your idea might be to study the impact of price changes on consumer behavior. A hypothesis might be that consumers will react positively to a change in packaging. A theory could be that consumers will respond positively to a change in packaging because it increases the product’s appeal to a wider audience.
Your research will determine if you prove or disprove the theme of your research. In turn, you’ll have some solid data to guide if the price change will reap the benefits that you envisioned.
You will want to decide the best research design for your study. Is there quantitative data available, like hard financial or statistical data? Or will you be interviewing subjects about their opinions and reactions?
Choose a research approach that fits your research strategy. Whether it is experimental, descriptive, explanatory, qualitative or quantitative, choose an approach that fits your study needs, timeline, and resources.
Which audience will give you the insights you need? Your research sample should reflect your ideal client, target audience, employee pool, or another group. Consider which demographics like age, gender, income data, and other data will give you objective results that can be generalized to a larger population.
Consider how many people you would like to reach. When it comes time for analysis, you will want to run significance testing to see if there are any meaningful differences between variables. For instance, you could cut your data by demographic variables, such as males and females, then run significance testing to see if there are any meaningful differences in your survey results between males and females. You will want to have at least n=100-200 respondents for each variable you plan to cut your data by. If you plan to cut your data by individual variables, you will need to have a larger sample size.
Does your target audience need to be studied over time in a longitudinal study? Or do you want a cross-sectional, one-point-in-time understanding from your audience?
If your research design lends itself to survey questions, you can design questions that will give you the responses that will prove or disprove your study’s hypothesis.
Survey firms offer test banks of questions that you can easily choose from to begin your study. They will help you collect answers to multiple choice, yes/no, or open-ended answers that reveal insights into your target audience’s behavior.
Whether you are using surveys, interviews, or another approach, you have now arrived at the most critical part of your study.
Collect your data carefully. Consider creating databases, filing systems, sorting processes, and other methods to capture your data. Eventually, you will use a statistical approach that matches your research design to give you the reports you need for analysis.
If you are using free survey templates, many reports come with your survey, along with powerful analytics that make it easy for your team to interpret. You’ll be able to compare them to your study’s goal and reach a conclusion about your theory or hypothesis.
Don’t forget to make your results presentation-ready to show to your executive sponsor, client, or other stakeholders.
Design is indeed good business when it comes to effective market research. Taking the time to choose the best design, and then taking the steps to execute on that approach will help assure statistically valid results that can guide better decision-making.
SurveyMonkey is a leader in helping companies design surveys that reach their research design goals. No matter what approach you use, we can help you create a survey that clarifies your research goal and saves you time with expert-written questions.
Collect market research data by sending your survey to a representative sample
Get help with your market research project by working with our expert research team
Test creative or product concepts using an automated approach to analysis and reporting