The ultimate guide to research design for business

If you have a business problem that you’re trying to solve — from product usage to customer engagement — doing research is a great way to understand what is going wrong.

Yet despite this, less than 40% of marketers use consumer research to drive decisions^[1].

So why are businesses missing out on vital business insights that could help their bottom line?

One reason is that many simply don’t know which research method to use to correctly investigate their problem and uncover insights.

This is where our ultimate guide to research design can help. But first…

What is research design?

Research design is the overall strategy (or research methodology) used to carry out a study. It defines the framework and plan to tackle established problems and/or questions through the collection, interpretation, analysis, and discussion of data.

While there are several types of research design (more on that later), the research problem defines which should be used — not the other way around. In working this way, researchers can be certain that their methods match their aims — and that they’re capturing useful and actionable data.

For example, you might want to know why sales are falling for a specific product. You already have your context and other research questions to help uncover further insights. So, you start with your research problem (or problem statement) and choose an approach to get the information you need.

Download our free eBook: How to get inclusive research design right

Key considerations before a research project

After you have your research problem and research questions to find out more information, you should always consider the following elements:

• Do you want to use a qualitative or quantitative approach?
• What type of research would you like to do (e.g. — create a survey or conduct telephone interviews)?
• How will you choose your population and sample size fairly?
• How will you choose a method to collect the data for ease of operation? The research tool you use will determine the validity of your study
• How will you analyse data after collection to help the business concern?
• How will you ensure your research is free from bias and neutral?
• What’s your timeline?
• In what setting will you conduct your research study?
• Are there any challenges or objections to conducting your research — and if so, how can you address them?

Ultimately, the data received should be unambiguous, so that the analysts can find accurate and trustworthy insights to act upon. Neutrality is key!

Types of approaches in research design

There are two main approaches to research design that we’ll explore in more detail — quantitative and qualitative.

Qualitative research design

Qualitative research designs tend to be more flexible and inductive (broad generalisations rather than specific observations), allowing you to adjust your approach based on the information you find throughout the research process. It looks at customer or prospect data (X data).

For example, if you want to generate new ideas for content campaigns, a qualitative approach would make the most sense. You can use this approach to find out more about what your audience would like to see, the particular challenges they are facing (from a business perspective), their overall experiences, and if any topics are under-researched.

To put it simply, qualitative research design looks at the whys and hows — as well as participants’ thoughts, feelings, and beliefs. It seeks to find reasons to explain decisions using the data captured.

However, as the data collected from qualitative research is typically written rather than numerical, it can be difficult to quantify information using statistical techniques.

When should you use qualitative research design?

It is best used when you want to conduct a detailed investigation of a topic to understand a holistic view. For example, to understand cultural differences in society, qualitative research design would create a research plan that allowed as many people from different cultures to participate and provided space for elaboration and anecdotal evidence.

If you want to incorporate a qualitative research design, you may choose to use methods like semi-structured focus groups, surveys with open-ended questions, or in-depth interviews in person or on the phone.

Quantitative research design

Quantitative research design looks at data that helps answer the key questions beginning with ‘Who’, ‘How’, ‘How many’ and ‘What’. This can include business data that explores operation statistics and sales records and quantifiable data on preferences.

Unlike qualitative research design, quantitative research design can be more controlled and fixed. It establishes variables, hypotheses, and correlations and tests participants against this knowledge. The aim is to explore the numerical data and understand its value against other sets of data, providing us with a data-driven way to measure the level of something.

When should you use quantitative research design?

If you want to quantify attitudes, opinions, behaviours, or any other defined variable (and general results from a large sample population), a quantitative approach is a way to go.

You could use quantitative research to validate findings from qualitative research. One provides depth and insight into the whys and hows, while the other delivers data to support them.

If you want to incorporate a quantitative research design, you may choose to use methods like secondary research collection or surveys with closed-ended questions.

Now that you know the differences between the two research approaches (though you can find out more), we can go further and address their sub-categories.

Research methods: the subsets of qualitative and quantitative research

Depending on the aim/objective of your research, there are several research methods (for both qualitative and quantitative research) for you to choose from:

Types of quantitative research design:

• Descriptive – provides information on the current state of affairs, by observing participants in a natural situation
• Experimental – provides causal relationship information between variables within a controlled situation
• Quasi-experimental – attempts to build a cause and effect relationship between an independent variable and a dependent variable
• Correlational – as the name suggests, correlational design allows the researcher to establish some kind of relation between two closely related topics or variables

Types of qualitative research design:

• Case studies – a detailed study of a specific subject (place, event, organization)
• Ethnographic research – in-depth observational studies of people in their natural environment (this research aims to understand the cultures, challenges, motivations and settings of those involved)
• Grounded theory – collecting rich data on a topic of interest and developing theories inductively
• Phenomenology – investigating a phenomenon or event by describing and interpreting the shared experiences of participants
• Narrative research – examining how stories are told to understand how participants perceive and make sense of their experiences

Other subsets of qualitative and quantitative research design

• Exploratory – explores a new subject area by taking a holistic viewpoint and gathering foundational insights
• Cross-sectional – provides a snapshot of a moment in time to reflect the state
• Longitudinal – provides several snapshots of the same sample over a period to understand causal relationships
• Mixed methods – provide a bespoke application of design subsets to create more precise and nuanced results
• Observational – involves observing participants’ ongoing behavior in a natural situation

Let’s talk about these research methods in more detail.

Experimental

As a subset of quantitative research design types, experimental research design aims to control variables in an experiment to test a hypothesis. Researchers will alter one of the variables to see how it affects the others.

Experimental research design provides an understanding of the causal relationships between two variables – which variable impacts the other, to what extent they are affected, and how consistent is the effect if the experiment is repeated.

To incorporate experimental research design, researchers create an artificial environment to more easily control the variables affecting participants. This can include creating two groups of participants – one acting as a control group to provide normal data readings, and another that has a variable altered. Therefore, having representative and random groups of participants can give better results to compare.

Image source: World Bank Blogs

Descriptive

Descriptive research design is a subset of qualitative design research and, unlike experimental design research, it provides descriptive insights on participants by observing participants in an uncontrolled, geographically-bound natural environment.

This type gives information on the current state of participants when faced with variables or changing circumstances. It helps answer who, what, when, where, and how questions on behaviour, but it can’t provide a clear understanding of the why.

To incorporate a descriptive research design, researchers create situations where observation of participants can happen without notice. In capturing the information, researchers can analyse data to understand the different variables at play or find additional research areas to investigate.

Exploratory

Exploratory research design aims to investigate an area where little is known about the subject and there are no prior examples to draw insight from. Researchers want to gain insights into the foundational data (who, what, when, where, and how) and the deeper level data (the why).

Therefore, an exploratory research design is flexible and a subset of both quantitative and qualitative research design.

Like descriptive research design, this type of research method is used at the beginning stages of research to get a broader view, before proceeding with further research.

To incorporate exploratory research design, researchers will use several methods to gain the right data. These can include focus groups, surveys, interviews in person or on the phone, secondary desk research, controlled experiments, and observation in a natural setting.

Cross-sectional

Just like slicing through a tomato gives us a slice of the whole fruit, cross-sectional research design gives us a slice representing a specific point in time. Researchers can observe different groups at the same time to discover what makes the participant behaviour different from one another and how behaviour correlates. This is then used to form assumptions that can be further tested.

There are two types to consider. In descriptive cross-sectional research design, researchers do not get involved or influence the participants through any controls, so this research design type is a subset of quantitative research design. Researchers will use methods that provide a descriptive (who, what, when, where, and how) understanding of the cross-section. This can be done by survey or observation, though researcher bias can be an undesirable outcome if the method is not conscious of this.

Analytical cross-sectional research design looks at the why behind the outcome found in the cross-section, aligning this as a subset of qualitative research design. This understanding can be gained through emailed surveys. To gain stronger insights, group sample selection can be altered from a random selection of participants to researchers selecting participants into groups based on their differences.

Since only one cross-section is taken, this can be a cheaper and quicker way to carry out research when resources are limited. Yet, no causal relationships can be gained by comparing data across time, unlike longitudinal research design.

Longitudinal

Longitudinal research design takes multiple measures from the same participants or groups over an extended period. These repeated observations enable researchers to track variables, identify correlations and see if there are causal relationships that can confirm hypothesis predictions.

As the research design is focused on understanding the why behind the data, this is a subset of qualitative research design. However, the real-time data collection at each point in time will also require analysis based on the quantitative markers found through quantitative research design.

Researchers can incorporate longitudinal research design by using methods like panel studies for collecting primary data first-hand. The study can be retrospective (based on event data that has already occurred) or prospective (based on event data that is yet to happen).

While being the most useful method to get the data you need to address your business concern, this can be time-consuming and there can be issues with maintaining the integrity of the sample over time. Alternatively, you can use existing data sets to provide historical trends (which could be verified through a cross-sectional research design).

Mixed methods

Mixed methods aim to provide an advanced and bespoke response to solving your business problem. It combines the methods and subsets above to create a tailored method that gives researchers flexibility and options for carrying out research.

The mixed-method research design gives a thorough holistic view of the layers of data through quantitative and qualitative subset design methods. The resulting data is strengthened by the application of context and scale (quantitative) in alignment with the meaning behind behaviour (qualitative), giving a richer picture of participants.

Mixed method research design is useful for getting greater ‘texture’ to your data, resulting in precise and meaningful information for analysis. The disadvantages and boundaries of a single subset can be offset by the benefits of using another to complement the investigation.

This subset does place more responsibility on the researcher to apply the subset designs appropriately to gain the right information. The data is interpreted and assessed by the researcher for its validity to the end results, so there is potential for researcher bias if they miss out on vital information that skews results.

Image Source: Full Stack Researcher

Find the research design method(s) that work for you

No matter what information you want to find out — there’s a research design method that’s right for you.

However, it’s up to you to determine which of the methods above are the most viable and can deliver the insight you need. Remember, each research method has its advantages and disadvantages.

It’s also important to bear in mind (at all times), the key considerations before your research project:

• Do you want to use a qualitative or quantitative approach?
• What type of research would you like to do (e.g. — create a survey or conduct telephone interviews)?
• How will you choose your population and sample size fairly?
• How will you choose a method to collect the data for ease of operation? The research tool you use will determine the validity of your study
• How will you analyse data after collection to help the business concern?
• How will you ensure your research is free from bias and neutral?
• What’s your timeline?
• In what setting will you conduct your research study?
• Are there any challenges or objections to conducting your research — and if so, how can you address them?.

But if you’re unsure about where to begin, start by answering these questions with our decision tree:

Image Source: Research Gate

If you need more help, why not try speaking to one of our Qualtrics team members?

Our team of experts can help you with all your market research needs — from designing your study and finding respondents, to fielding it and reporting on the results.

^[1] https://www.thinkwithgoogle.com/consumer-insights/consumer-trends/marketing-consumer-research-statistics/