How to Write a Research Question & Hypothesis (with Examples)
A good research paper starts with a good question, specifically a good research question.
Good research questions help us develop strong hypotheses, or predictions, for our papers. Writing a hypothesis has things in common with writing a master's thesis, only we aren't making a claim. Rather, we're making a prediction.
If you feel a little lost about the relationship between research questions and hypotheses, worry not. In this post, we'll dive into writing a good research question and hypothesis, and we'll show examples of good research questions.
Research Question vs. Hypothesis: What Is the Difference?
A research question is based on research you've gathered on a certain topic. A hypothesis, on the other hand, aims to answer your research question with a testable prediction.
At their simplest, research questions and hypotheses work together to build the foundation of your research paper. Your research question helps you hone your research, while your hypothesis tells you what to test and eventually write.
Think about building a house from the ground up. Your research question is like the ground on which you build the house, and your hypothesis is like the first floor.
You can't build a house without the ground and the first floor, and you can't build a research paper without a research question and a hypothesis.
What Is a Research Question?
A research question states what the content of your research paper will aim to answer. It describes a problem to be addressed through testing and analysis, and it should be solved by the paper's conclusion.
Research questions address a specific aspect of a specific topic. You can change and refine your research question throughout the research process as you investigate the topic and learn more about it.
Typically, one research question is enough to guide your paper, but some larger research projects may require multiple research questions.
What Makes a Good Research Question?
Writing a research question seems straightforward, but what makes a good one?
Before you start writing, consider the criteria listed below. You should be sure to meet all of them to ensure that your research question will work well.
Examples of Good and Bad Research Questions
Good Research Questions
Now that we know what makes a strong research question, let's look at a few examples of them in the social and natural sciences.
What impact has the affordability of housing in Los Angeles had on homelessness in the last 20 years?
What is the effect of pollen on seed viability and fruit quality?
How does the reproduction rate of brown recluse spiders change by season in North America?
Not-So-Good Research Questions
Now, let's take a look at some examples of less strong research questions, so you understand what to avoid and why.
Which city deals with homelessness better: Los Angeles or New York?
The issue with this question is that we don't have a clear definition of "better." Remember, your research question needs to be researchable. You'd need to develop criteria for what is "better" to make this one work.
What effect does pollen have on fruit?
In this case, we need more specificity about the effect of pollen on fruit, because it likely affects fruit in many ways. What specific effects are we studying? This could relate to properties such as the overall quality of the fruit and/or its seed viability.
What factors lead to low reproduction rates in brown recluse spiders?
This question is too broad because there might be many different factors to consider. Choose one factor to focus on in your research. This will narrow down your research scope considerably and help strengthen your research question.
How to Write a Research Question
A strong research paper starts with a specific question that defines the boundaries and limits of your study. It's like putting a box around the issue or problem you're studying so that your research is clear and contained.
Without a research question, you run the risk of drawing incorrect conclusions about your topic. Worse, you might leave out critical information.
Follow the steps below to write your research question.
Step 1: Determine the Requirements of Your Assignment
When you're just starting out, be sure to follow the requirements of your assignment. Writing the best research question will depend on what your professor has assigned you.
Questions to ask yourself at this stage: Is there a specific format I am supposed to follow? Is there a formula I should use?
Step 2: Choose a Topic
The next step is to choose a research topic. We suggest starting with something broad. Explore the subject areas that interest you. This makes research a lot more fun!
When you start with a broad topic, you're much likelier to find unanswered questions in your field of interest. You'll also start to notice different subtopics that relate to your general topic.
Questions to ask yourself at this stage: What do I find useful or interesting about this topic? Are any sources raising good questions?
Step 3: Preliminary Research
Now that you have a broad topic in mind, you can start conducting the research necessary to develop your research question.
When it comes to conducting your research, use sources that are current, relevant, and unbiased. Try skimming resources at your university's library, such as newspapers, academic journals, encyclopedias, or websites with .edu or .org as their domain.
When in doubt, stick to books or academic journals to uncover credible scholarly research. They are often the most trustworthy sources.
Questions to ask yourself at this stage: Do any gaps exist in the research? Are there any limitations to the existing research? Who is affected by this topic?
Step 4: Narrow Your Topic
Now, it's time to narrow down the broad topic chosen in step 2 to a specific one. Failing to define your broad topic further is one of the biggest mistakes you can make when developing researchable questions.
The good news is that your research will help you discover subtopics that are worthy of exploration.
For example, let's say you wanted to investigate the topic of traffic in crowded cities. This topic alone is much too broad. What, in particular, about traffic in crowded cities can we explore?
The topic of traffic is interesting. For example, why do traffic jams happen? What do people think is the cause of traffic jams but actually isn't?
You could narrow down your topic simply by delving into the specifics of one of these questions. The gradual process of refining this topic could look like this:
Traffic in crowded cities > Causes of traffic in crowded cities > Non-accident-related traffic jams > Non-accident-related traffic jams in Los Angeles
You might also consider what effect traffic has on society at large, as relevance is one of the keys to a great research question.
Questions to ask yourself at this stage: Is my topic narrow enough to be covered within the scope of my research? Would a further subtopic or more specific angle be more appropriate?
Step 5: Write Your Research Question
Now it's time to write your research question. In simple terms, this means asking a question about the research you've conducted. Continuing with the traffic jam example given in the previous section, let's write a sample research question.
If we wanted to know how non-accident-related traffic jams in Los Angeles affect society, we might start with the following question:
What effect do non-accident-related traffic jams have on society in Los Angeles?
This is a good start, but we might also refine this further by specifying one aspect of Los Angeles's society.
What effect do non-accident-related traffic jams have on regional economic health in Los Angeles?
Even greater specificity can help further refine most topics. For example, what region(s) in Los Angeles will be studied? How is "economic health" defined in the scope of this paper?
Questions to ask yourself at this stage: What do I want to explore after my preliminary research? Did I narrow my topic down enough?
Step 6: Evaluate the Question
Now it's time to evaluate your research question to determine whether it's strong enough.
Questions to ask yourself at this stage:
- Is my research question focused? Make sure it focuses on one problem or concern.
- Are primary and secondary sources on this topic available? Primary sources provide first-hand evidence (e.g., interviews you have conducted), and secondary sources offer second-hand information (e.g., reviews you have read).
- Is this research question feasible? Do I have enough time and resources to investigate the answer? If not, you might consider narrowing the research question.
- Is my research question specific? Always define who, what, where, and when.
- Is my research question complex? Make sure the answer to the research question is not simply "yes" or "no."
- Is my research question relevant? Is it related to other research in my field? Does it identify a gap in the current knowledge on the topic?
If you need help developing an effective research question, try Merudio's Early Draft Editing service.
Let's Develop Your Research Question
What Is a Hypothesis?
A hypothesis is a testable prediction you make at the start of your research process. It describes your expectations regarding a specific circumstance or a relationship between two or more variables. Essentially, it acts as an untested answer to your research question.
Depending on the type of testing you're doing, your hypothesis could include null and alternative hypotheses.
A null hypothesis predicts that there is no relationship between the two variables you're testing, while an alternative hypothesis predicts that there is a relationship.
What Is the Difference between a Hypothesis and a Theory?
Unlike a hypothesis, a theory has been rigorously studied and is often widely accepted by scientists or other members of the academic community.
A theory explains an event or phenomenon and is backed by data. A hypothesis, in contrast, is a prediction you make about the relationship between two variables before testing it.
Because theories have undergone rigorous experimentation in the scientific community, the probability that a theory is accurate is higher than that of a hypothesis.
What Is an Alternative Hypothesis?
We've already defined a hypothesis as a testable prediction proposed at the beginning of your paper. However, some testing, such as statistical testing, requires that you restate your hypothesis as an alternative and a null hypothesis.
An alternative hypothesis predicts a relationship or association between two variables. This is the opposite of a null hypothesis, which predicts that no relationship exists between two variables. Together, the alternative and null hypotheses take the place of a simple hypothesis.
What Is a Null Hypothesis?
A null hypothesis predicts that no relationship exists between two variables.
Let's say your research question is to determine whether a relationship exists between how often a dog barks and the dog's size.
If you develop a hypothesis that predicts that small dogs tend to bark more than large dogs, the null hypothesis is as follows:
(H0) Small dogs, on average, do not bark more than large dogs.
When answering your research question, you would restate your prediction (or hypothesis) as a null hypothesis (written as H0) and an alternative hypothesis (written as Ha or H1), so it can be mathematically tested.
An alternative hypothesis describes a predicted association or effect between two variables. It opposes the null hypothesis.
The alternative hypothesis to the null hypothesis in the previous section is as follows:
(Ha) Small dogs, on average, bark more than large dogs.
Here, we predict an association between the frequency of barking and the smallness of dogs. This is the opposite of what we predicted in the null hypothesis, which was that small dogs do not bark more than large dogs.
A good way to remember the definition of an alternative hypothesis is that it is the opposite of, or alternative to, a null hypothesis.
There are two types of alternative hypotheses: directional and nondirectional.
1. Directional Hypothesis
A directional hypothesis relates to the directional change expected when your hypothesis is tested. If your hypothesis predicts that the change will move in a positive or negative direction, you have a directional hypothesis.
For example, if we predicted that small dogs bark more than large dogs, the frequency of barking would be expected to move in a negative direction as a function of the size of dogs.
Because the predicted change is moving in a certain direction, we have a directional hypothesis.
2. Nondirectional Hypothesis
A nondirectional hypothesis relates to the change expected to happen when your hypothesis is tested. It applies when the direction of change isn't specified.
For example, if we predicted that the frequency of barking in small and large dogs would be different but didn't specify whether it would be higher or lower, we would have a nondirectional hypothesis. This is because the direction of change would be neither positive nor negative.
A null hypothesis predicts that there is no association or effect between two variables. This is the opposite of the alternative hypothesis.
To write a null hypothesis, ask yourself what your prediction would look like if there were no changes in the two variables you are testing.
For example, in our hypothesis that small dogs bark more than large dogs, our null hypothesis would predict that a dog's size has nothing to do with barking frequency.
(H0) Small dogs, on average, do not bark more than large dogs.
Now, let's look at some examples of alternative (directional/nondirectional) and null hypothesis statements for various research questions.
1) What effect does the daily use of Instagram have on suicide rates in adolescents aged 13–17?
Null hypothesis: The daily use of Instagram has no effect on suicide rates in adolescents aged 13–17.
Directional hypothesis: The daily use of Instagram causes suicide rates to increase in adolescents aged 13–17.
Nondirectional hypothesis: The daily use of Instagram affects suicide rates in adolescents aged 13–17.
2) What effect does type 2 diabetes have on the lives of American men aged 45–60?
Null hypothesis: Type 2 diabetes has no effect on the lives of American men aged 45–60.
Directional hypothesis: Type 2 diabetes has a negative effect on the lives of American men aged 45–60.
Nondirectional hypothesis: Type 2 diabetes has some effect on the lives of American men aged 45–60.
3) What impact has the affordability of housing in Los Angeles had on homelessness in the last 20 years?
Null hypothesis: The affordability of housing in Los Angeles has had no impact on homelessness in the last 20 years.
Directional hypothesis: The affordability of housing in Los Angeles has had a negative impact on homelessness in the last 20 years.
Nondirectional hypothesis: The affordability of housing in Los Angeles has impacted homelessness in the last 20 years.
4) What is the effect of pollen on seed viability and fruit quality?
Null hypothesis: Pollen has no effect on seed viability or fruit quality.
Directional hypothesis: Pollen has a positive effect on seed viability and fruit quality.
Nondirectional hypothesis: Pollen affects seed viability and fruit quality.
5) How does the reproduction rate of brown recluse spiders change with seasonal changes in North America?
Null hypothesis: There is no change in the reproduction rate of brown recluse spiders with seasonal changes in North America.
Directional hypothesis: The reproduction rate of brown recluse spiders increases as the season in North America changes from winter to spring.
Nondirectional hypothesis: There is some difference in the reproduction rate of brown recluse spiders with seasonal changes in North America.
How to Write a Hypothesis
Your hypothesis is an untested prediction about the answer to your research question. After you've done some preliminary research on your research question, you should have enough information to write your hypothesis.
Keep in mind that your hypothesis should do the following:
- Predict cause and effect
- Be testable
- Contain variables
- Be clear and simple
- Make a statement rather than ask a question
Try not to worry about drafting the perfect hypothesis the first time. Hypotheses can be refined to ensure that they are clear and include relevant variables.
Use the tips below to help you develop a strong hypothesis.
Step 1: Start Investigating Your Research Question
The first step in writing your hypothesis is to do some preliminary investigation on your research question.
If your research question asks, "What effect does type 2 diabetes have on the lives of American men aged 45–60?" then your research should center on finding an answer to that question.
Related: Check out our list of free research databases.
Step 2: Answer Your Research Question
Next, use your research to come up with an answer to your research question.
What educated assumptions can you make about what you found while researching? Were you able to identify two variables to study?
These are important questions to ask when formulating your hypothesis.
Step 3: Draft a Hypothesis
At this point, you'll have some idea of your hypothesis based on your research question.
Sometimes, it helps to write a clear statement about what you predict will happen in your test.
Sample hypothesis: Type 2 diabetes affects the lives of American men.
Step 4: Revise Your Hypothesis
The next step in writing a hypothesis is to revise it. Your hypothesis forms the basis for the validity of your research, so you want to make sure it's strong. Verify that it has two variables and that it's testable and specific.
Your hypothesis should also predict cause and effect. One way to verify this is by using the words "if" and "then." For example, if American men aged 45–60 have type 2 diabetes, then it will have a negative effect on their lives.
Step 5: Create a Null Hypothesis
The last step is to create a null hypothesis, which states that there will be no effect on or change in a variable. Creating a null hypothesis gives us more information about our test.
In the case of our sample research question, the null hypothesis is as follows: Type 2 diabetes has no effect on the lives of American men aged 45–60.
Criteria for a Good Hypothesis
You won't always know a good hypothesis from a bad one when you're starting out. It takes time to develop a muscle for academic writing. The more you practice, the better you'll get.
If you're ever unsure whether your hypothesis will hold up, use the questions below to check its strength:
Use these questions as a checklist to refine your hypothesis and make it stronger. They may help you identify any weaknesses you didn't see before, and you'll be a better researcher because of it!
A good research paper begins with the identification of a strong research question. This sets up a sturdy foundation for testing and experimenting.
From there, a strong hypothesis establishes a prediction about what the research will tell you. This prediction guides the remainder of your writing.
To ensure your research question and hypothesis hold up, try Merudio's Early Draft Editing service.
Let's Develop Your Research Question
We will continue to update Merudio; if you have any questions or suggestions, please contact us!
Follow us on LinkedIn, Twitter, and Facebook.