Research design is the framework for your research. It gives you a plan to follow to achieve your research goals. There are different types of research designs. These varieties are differentiated by their data collection and data analysis methods.
Research design is a blueprint for your research. This article describes the different research designs and their purpose. These descriptions will help you choose the best design for your study.
These study designs can either be quantitative research or qualitative research designs. Below are the main research methods used in social sciences and scientific studies:
Longitudinal research follows the same subjects over time. Therefore, you only have one group of subjects instead of two different groups.
For example, you are studying the effects of a newly implemented education program. Each student has participated in this new education program. After one year, you assess their understanding of the program. Then, after two years, you re-evaluate their understanding.
You would compare this data with a group that has not experienced the education program. This comparison will allow you to identify the effects of the education program on student performance.
This research design is suitable for assessing the impact of an intervention. It can also help you understand how people change over time.
The main drawback of the longitudinal design is that it takes a long time to collect and analyze data.
- Longitudinal design measures change over time.
- It is good for measuring the impact of interventions.
- A longitudinal study requires a large sample size.
- This design takes a long time to complete the study. It is not suitable for experiments with short exposure.
With this research design, you compare two or more groups at one time. Each group falls under the same category.
For example, you are studying how exercise affects people of different ages.
You would need to collect data from two different groups of people. One group of subjects is teenagers, while the other group consists of middle-aged adults.
This research design is suitable for identifying differences between groups of participants, but it cannot identify the cause of these differences.
This design is common in social psychology research studies.
- You can compare different groups at one time.
- It is suitable for identifying differences between groups of participants.
- It cannot determine cause and effect.
- It cannot tell whether change over time is due to aging or other factors.
This type of research design focuses on individual cases or individuals instead of focusing on group comparisons. It allows you to determine the relationship between specific risk factors and diseases.
For example, you are studying the effects of smoking on lung cancer. Instead of comparing people who smoke and don’t smoke, you would identify two groups of people.
One group consists of individuals with lung cancer. The second group includes people without lung cancer.
You would then collect information about how many years they have smoked and other risk factors for lung cancer.
- It allows you to determine the relationship between specific risk factors and diseases.
- Case-control design is great for studying risk factors and causes of diseases.
- It only focuses on specific cases, so it only investigates the differences between the groups, not why these differences exist.
- Case-control design can lead to selection and recall bias.
Researchers use the case study design for studying individual cases or events. This type of research design is common in educational fields and the social sciences.
For example, you are studying a particular school’s education program. To conduct the study, you would collect information from teachers and students at this school. You would analyze these results and compare them to the results from another school.
- This method is good for studying events or individuals in great detail. Its results are usually limited to these specific cases, however.
- Results are usually limited to specific cases.
- The sample collection for this method is more complicated than the other methods. It requires you to make sure that your sample represents all people or events of interest.
This research design describes the characteristics of a group or population. It allows you to identify the distribution and frequency of variables in your particular sample.
For example, you are studying people who live in different states. You would need to find out where each participant lives and which age they are.
- This study design gives a vivid description of your population.
- Results cannot be generalized to other populations.
- It does not allow inferences about other populations.
- It does not allow you to compare your findings to the larger population.
This research design works best when determining cause and effect (how one variable affects another). It is the most common research design used in psychology.
This method compares two or more groups to determine which group performs better on a particular task.
You would need to divide your participants at random into different groups, each of which receives a different treatment or intervention. In the end, you would compare the groups’ performance on some tasks.
The experimental design worked best in testing the cause and effect relationship between variables. The independent variable is the group that receives different treatments. The dependent variable is how this group performs on a particular task or measure.
- This research design is suitable for determining cause and effect.
- Results can be generalized to other populations.
- It can be time-consuming and costly to implement.
- Results may only represent the specific participants you test, not other populations or individuals.
- It only generates results for groups of people and not single individuals
The mixed-methods design combines elements of both experimental and descriptive research designs.
For example, you are studying high school students’ grades in math class. You would compare different groups of students who are taught math differently.
You would then compare your research findings in this study with results from other studies. It is applied when the same research question is answered in two different ways.
- Results can be generalized to other populations.
- This design is suitable for studying difficult situations to test, such as people’s behavior or experiences. It can also address ethical concerns.
- This design is more complicated to implement than other methods.
- Researchers also have to make sure that the two investigations match up perfectly.
The causal relationship is a study research design used to find out if one variable causes another.
You would need to find two groups of people; the experimental and the control groups. Look for differences in the groups’ performance on a particular task.
For example, you need to divide your participants into two groups: one group would perform some tasks, and the other group would not. You would then observe how well both groups did on this task.
If the group that performs the task does better, you will assume that performing the task caused their success.
It is best used for studying situations where it is difficult to manipulate the independent variable.
The experimental design was best for testing cause-and-effect relationships between variables.
The independent variable was the group that received different treatments. The dependent variable was the group’s performance on a particular task or measure.
- This research design is suitable for studying situations that are difficult to test. For example, many people’s experiences or behavior are impossible to test in an experimental setting. It can also often provide more information than other methods.
- It is best used for studying situations where it is difficult to manipulate the independent variable.
- The experimental design is best for testing cause-and-effect relationships between variables. The independent variable was the group that received different treatments. The dependent variable was the group’s performance on a particular task or measure.
- Results can only be generalized to specific groups of people, not all populations or individuals.
- This design has some biases. For instance, researchers may instruct two groups of students to act differently on their tests so that one group performs better. This is unethical since it leaves a bad feeling on those who failed.
- It is usually more expensive and time-consuming to implement than other research designs.
This research design looks at groups of people over time. One advantage of this design is that you are studying real-world changes over time. It is also helpful for studying changes that can take a long time to occur.
The sampling process is where this design gets complicated. You need to select a group of people (cohort), and then you would look at them repeatedly over time.
You would need to wait until the desired outcome has occurred for as long as it takes. Cohort designs can be expensive and time-consuming to implement and might not be practical.
- Cohort design is suitable for studying events that are difficult to quantify.
- It can provide information about long-term changes in people’s behavior and experiences.
- The design is great in situations that are impossible to test, such as people’s behavior or experiences.
- This research design is expensive and time-consuming because you need to wait until the desired outcome has occurred for as long as it takes.
- It may not be practical or economically feasible. It can be challenging to measure the desired outcome.
- The cohort design was best for studying long-term changes in people’s behavior and experiences over time. It was not practical, informative, or ethical because it was expensive and time-consuming, and you need to wait until the desired outcome has occurred for as long as it takes.
It is a statistical procedure where the results from many studies are combined and analyzed together. The results are then re-analyzed to find out if the results are consistent.
This design works best when making conclusive statements about particular research questions.
- Meta-analysis is good for making conclusive statements about particular research questions.
- You can use this design to make statistical corrections when designing new studies.
- It can be challenging to combine the differences between studies and implement the analysis correctly.
- This design is not suitable for making recommendations about interventions or treatments.
- The meta-analysis was best for making conclusive statements about particular research questions. However, the design is not good for making recommendations about interventions or treatments.
This research design is similar in many ways to the experimental design. It can still help you determine if one variable causes another, but the procedures are not as tightly controlled.
This design is applicable in studying situations where it would be difficult or unethical to manipulate the independent variable.
The main difference between the experimental and quasi-experimental designs is that; some participants are not randomly assigned to groups in this design.
The sample size has to be larger in this design due to the lack of experimental control. This makes it less likely that the research will produce a significant result.
- This design is incredible for studying situations where it is hard to manipulate the independent variable. E.g., People’s behavior and experiences.
- It can still help you determine if one variable causes another.
- The sample size has to be larger in this design due to the lack of experimental control, which makes it less likely that the research will produce a significant result.
- The quasi-experimental design is not as tightly controlled as the experimental design.
- It can be harder to identify the cause of a change.
- This design is limited. It should only be applied when it is impossible to manipulate the independent variable. E.g., studying long-term changes in people’s behavior and experiences over time.
This research design is used to determine the relationship between variables. This involves measuring both dependent and independent variables.
You would compare the changes in the two variables to determine whether there is a relationship between them.
For example, you might want to see if people who drink more coffee are more likely to get cancer. The results of this design could be due to the variables both being caused by a third, unmeasured variable.
The three types of correlations are:
- Positive correlation: The variables move in the same direction. One variable likely causes another if there is a positive correlation between them.
- Negative correlation: The variables move in opposite directions. One variable causes the other if there is a negative correlation between them.
- No correlation: There does not appear to be any correlation between the variables.
- This design can give you information about how much impact changing one variable will have on the other.
- It is great for determining if there is a statistical relationship between variables.
- Correlation design doesn’t determine causation.
- The results of this design could be due to the variables both being caused by a third, unmeasured variable.
This research design is a type of experiment that occurs in the natural setting or ‘field.’ It can be difficult to implement. It is often used in medicine, social science, and the natural sciences.
One example is an experiment where researchers are testing the impact of an anti-smoking campaign. You will need to work with different groups of people and monitor their smoking habits over time.
In this smoking experiment, accurate results should reflect the impact of the anti-smoking campaigns. Determine the smoking rate before, during, and after the campaigns. This study requires a follow-up on the participants to understand their behaviors.
This design helps study complex situations where it would not be easy to control outside variables, such as people’s behavior. To get reliable results, the researcher needs effective systems in place. Having a team of researchers working on one experiment helps the situation too.
Social experiments are another type of experiment where the study takes place in a natural environment. These studies are conducted in dynamic settings, and the researcher has minimal or zero control over the variables.
- Field experiments are well suited for examining people’s behavior in response to complex situations.
- These experiments can be difficult to implement.
- It can be challenging to control outside variables in a natural setting.
- This type of study is not as well suited for studying cause and effect. This is because it cannot control all the possible consequences of being in the experiment
The exploratory design is suitable when the researcher has little idea of what variables might be related.
It can still help answer questions about relationships between variables, but there is a greater risk of producing false results.
It is an essential first step in research because it lets the researcher see what variables might be already related.
You can use this design when one or more variables correlate, but there is no evidence of how these variables work together.
Exploratory research helps explain the relationships between variables, but it does not prove causation. This means that you cannot use this research as a basis for planning further research.
- The exploratory design lets the researcher see what variables might already be related.
- It can help answer questions about relationships between different
- This design is less controlled, and it is possible to produce false results.
- It does not demonstrate causation.
This design is applicable where there are no studies to guide the creation of a product.
For example, you might be designing an interface with little knowledge about how people use it.
Often, the UX researcher makes assumptions about usability and discusses these with stakeholders to get their feedback.
For example, when designing a new phone, you might ask a group of people to use old phones and record their experiences. You might also run a few tests with paper prototypes.
UX research can help determine if usability issues need to be changed for the product to work better.
- UX research is good for finding usability issues quickly.
- It can be suitable for exploring new ideas or designing relatively simple products.
- UX research helps find usability issues with a design quickly, which is helpful for iterative design.
- It lets you test and get feedback on designs without investing in them.
- UX research is good for finding usability issues quickly, but iterative design can still take a long time.
- It is not appropriate for finding the best solution to a problem.
You can use philosophical design when you don’t know what variables are related. It is exploratory and aims to learn about the connections between different variables, not test a hypothesis.
Testing and learning is its primary goal, not finding the truth about what variables are related.
It involves making guesses about variables that are related, then testing these with different approaches.
For example, you might use philosophical design to establish theories about how variables are connected before starting a full investigation.
- Philosophical design helps develop and test theories early on during the research project.
- It is useful for generating new ideas.
- Philosophical design is not appropriate for finding the best solution to a problem.
- It can be challenging to know if your theories and ideas are correct.
- This design can be challenging to implement well because it is based on guesswork rather than solid evidence.
Informative design is great for informing stakeholders about new areas of research but not testing specific hypotheses. You can design and research without worrying too much about how it will affect the final product.
- Informative design is suitable for exploring new areas of research.
- It works well in running controlled experiments that are not connected to the final product.
- This design can be used to do preliminary research that helps set the context for other types of research.
- It can also help tell stakeholders about your research at an early stage.
- Informative design is not appropriate for testing specific hypotheses or finding the best solution to a problem.
- It can be challenging to know if your results are accurate.
- Informative research demonstrates relationships between variables, but it does not aim to find causation.
Controlled experiment design is used when testing one variable that causes another. For example:
- One group of people (the control group) does not receive the treatment.
- The second group of people (the experimental group) does receive the treatment.
You would then compare the results for both groups to see if there is a difference.
Another example is that you might want to test whether giving people written instructions before using an interface improves the experience.
- One group of people would use the interface with no instructions.
- The other group would receive written instructions before they used the interface.
You could then compare the results for both groups to see if there is any difference in their experience.
- Controlled experiments are helpful because the researcher has complete control over how participants behave.
- They can also be very detailed in their results. For example, you can measure the impact of each factor on an outcome.
- Because controlled experiments happen in specific conditions, it is hard to make inferences about the wider population. This means that you must be careful when generalizing results to your target audience.
This research design works well when testing one variable that does not cause another.
You can’t manipulate the independent variable because you want to observe how it naturally unfolds. You can only observe the ‘outcomes’ of the variables.
For example, you might want to see if a new stereo system will improve people’s musical abilities.
You will gather a group of participants but not ask them to use the stereo. Instead, give them a pre-determined music test and observe the results.
This research design helps gather information about people’s experiences without changing them.
You should only use this design if you are sure that the independent variable has no other impact on people or events in the environment.
Using observational design can allow you to:
- Study phenomena in their natural context without changing them.
- Avoid ethical issues, like asking people to test a new drug when there is no need.
- You can’t use observational design if you are observing the impact of one variable on another. For this research, you will need an experimental design.
This design works best when the main interest is to describe what has happened in the past.
For example, you might want to know how countries have responded to different treaties or how common items have evolved.
You would gather information about past events by looking at primary sources, such as government documents, newspapers, and textbooks.
This research design helps gather information about historical events without manipulating them.
- Historical design can provide information about events that are not easily accessible.
- The researcher can get a clear picture of historical events without any contamination from other variables.
- Historical design is not suitable for finding out how factors interact.
- You cannot control factors and events, making it difficult to determine their roles in the past.
Researchers use survey design when they want to gather information about opinions, beliefs, and attitudes.
A survey asks people questions that are designed to yield specific types of data. You can then use statistics to analyze the results.
Let’s say you want to know what people think about using video clips to support learning. In this case, you could design a survey that asks questions about their attitudes towards video.
You can then analyze the results to find out whether there is a connection between attitudes and demographics.
This research design helps gather information from groups of people. You need to have access to a large enough group to be able to get reliable data.
Survey designs can also efficiently gather information about current behaviors that link with other elements (such as attitudes, opinions, and beliefs).
You should only use this design if you are sure that the independent variable has no other impact on people’s decisions or behavior.
- You can control variables. You can design surveys to ask participants specific questions and gather data from a large scale of people.
- You can gather information about past events by looking at primary sources.
- You need a large enough group to be able to get reliable data.
- The time frame for this design is long-term. If the variables change, your results can become unreliable.
- You need to have access to primary sources of information.
There are various factors to consider when determining the most appropriate research design for your study: These are:
- The primary purpose of the research and what you hope to gain from it
- The level of control you have over the research process and the participants
- The level of control you need over the findings and how to apply the latter. For example, in a commercial setting, you need to know how traders operate. Also, find out if you can do anything to manage their operations for more reliable results.
- Choosing to use a qualitative or quantitative approach will also affect your choice of research design. It is common to have a mix of both in UX research.
- Your research problem or objective. E.g., research problem: ‘We want to improve the way customers book their travel on our website.’ In this case, drawing inspiration from past trends will help, hence the historical design.
- Your key question(s) and possible objectives/aims
- Any theoretical framework you want to explore through the study
- Your personal and professional experience in collecting and analyzing data
- The time you have for the study
- The resources and support available
Often, teams do not have the time or resources to conduct an elaborate research project before designing new features. In this case, it is crucial to do as much up-front research as possible. This means deciding on the design before you begin looking at participants and analyzing data.
The process of creating a study design typically involves some or all of the following steps.
This decision will depend on your research problem and what you want to achieve.
Decide whether you want to carry out a qualitative or quantitative study, as this will affect your research design. There are different designs for quantitative and qualitative studies hence the need to choose well.
Based on your research problem, think about the people or user groups you need to recruit. You will need to identify groups of potential participants who match your criteria.
Think about various ways of recruiting participants. You can get them from a specific population or draw a sample from a larger population.
Determine the required number of participants
Decide how many participants you need to get reliable results. You need to recruit the required number of people within your budget and the time you have.
Also, consider an adequate sample size so that you can get reliable and valid results from your study. When the sample is too small, it doesn’t represent the population well.
Draft a legal document for your participants
If you want to guarantee confidentiality and anonymity, it is usually good to ask participants to sign a consent form. This form could be part of a general document that asks for permission to participate in your research.
This document should include information about what will happen during the study. It should also address how the data management team will handle the participants’ information. It should also state that the participation is voluntary and that the participants can withdraw at any time.
Decide on your data collection methods
Decide what data collection methods are suitable for your study. These should be appropriate to the research design and participants.
For example, designing a children’s games app will require data collection methods from an online shopping app.
Decide how you will analyze the data from your study.
- Will you need to code it manually or automatically?
- Will all the information be analyzed or just a sample?
- Will you need to report your findings qualitatively or quantitatively?
Asking the same questions over and over again can get boring. Consider having a list of FAQs that you can quickly refer to when designing a study.
Data can be quantitative or qualitative, and the collection methods differ depending on the type of data collected.
Quantitative data collection methods include experiments, questionnaires, and interviews.
Experiments collect data from a controlled or manipulated setting which researchers have complete control over. Most experiments involve randomly assigning participants to two or more conditions.
Questionnaires collect data through asking questions, which the researcher then analyzes. Some questionnaires are self-reports, which involve participants answering questions in writing or on a recording. It can also be interviewer-administered, where an interviewer asks the questions and records the responses.
Interviews involve participants verbally giving their views on specific topics, issues or questions. Researchers obtain data by asking open-ended and closed questions, probing, and using probes.
Qualitative data collection methods collect snapshots of what people think at a specific moment in time.
Data is collected through observations which can involve watching, listening, or recording. It can also be collecting things like photographs, objects, and documents.
Researchers use a variety of research design methods to conduct a study. These designs range from simple observational studies to complex randomized controlled trials. The simple designs have no control groups, while the complex ones have multiple groups and follow-up analyses.
The best design will depend on your specific objectives for the study, and how much time you have available. The information provided here will help you choose the design that best suits your research needs.