Research Design: Research Questions

Many research questions in education come from observing something unexpected in the classroom, reading about a new method of instruction, or learning about a new tool. For example, you might find that a student has a unique explanation for a concept and want to know if it would help other students. Or you might have read about a metacognitive strategy and want to know if using that strategy would improve learning outcomes in your class. Your research methods will depend heavily on what type of research question you have.

There are three main types of research questions: descriptive, relational, and causal. A good research question in education explicitly identifies

  • the group that you are studying, such as online students, and
  • the variables you intend to manipulate or measure, such as the instructional strategy or student experience.

Variables in educational research are usually related to

  • a characteristic of the learner, such as age,
  • a behavior, such as time spent interacting with course materials or performance,
  • a perception, such as self-reported experiences or responses to a survey about self-efficacy.

Good research questions are also open-ended. They typically start with “how,” “what,” or “why” and cannot be answered with a simple yes or no. For example, the answer to “How do men and women act differently?” provides much more information than “Do men and women act differently?”

Descriptive Questions

Descriptive questions ask about the characteristics of variables or groups of people. Answers to these questions can describe differences within or between variables based on data collected, but they cannot make inferences about the relationships among variables. For example, if you were exploring gender differences in study time, you might find that the women in your research spent more time studying than men, but with descriptive research, you couldn’t say that women are likely to spend more time studying than men.

Correct Example 1: What is the racial representation of students in online classes?

Incorrect Example 1: How likely are members of different racial groups to take online or on-campus courses? This is a relational question.

Correct Example 2: How often do students of different grade quartiles post on the forum?

Incorrect Example 2: What is the relationship between the number of forum postings and grade? This is a relational question.

Relational Questions

Relational questions ask about the relationships among variables but do not ask about the cause and effect between variables. Relational questions often ask if a change in one variable is related to a change in another variable. For example, a relational question might ask how the number of lectures missed is related to course grade. People often equate a relationship between variables with a causal relationship, which says that a change in one variable causes a change in the other. However, correlation does not equal causation for reasons detailed in a later post.

Correct Example 1: What is the relationship between college students’ gender and academic major?

Incorrect Example 1: How does students’ gender affect their choice of academic major? This is not possible to scientifically answer for reasons you’ll see in the “causal” section.

Correct Example 2: How does studying time relate to course grade in introductory programming courses?

Incorrect Example 2: How does changing students’ study time affect course grades? This is a causal question.

Causal Questions

Causal questions ask about the cause and effect relationship among variables. These types of questions are the most restrictive of the three because they require the researcher to manipulate the variable predicted to examine its effect. Sometimes researchers would like to find a causal relationship between variables but cannot manipulate a variable for practical or ethical reasons. For example, you cannot assign students to a gender for practical reasons nor assign people to an academic major for ethical reasons; therefore, you cannot determine that gender or major causes something else.

Correct Example 1: How does lecture medium (live or recorded) affect math students’ learning?

(Semi) Incorrect Example 1: How does course medium (online or on-campus) affect students’ learning? This question would likely not be causal because you usually cannot manipulate whether students take a course online or on-campus. It would be a correct causal question if you assigned students to online or on-campus courses.

Correct Example 2: What are the differences in learning outcomes between providing immediate or delayed feedback on homework?

Incorrect Example 2: What are the differences in learning outcomes between providing grades throughout the semester or providing only a final grade? This question could likely not be tested ethically.

Type of QuestionDefinitionExampleExplanation
DescriptiveAsks about characteristics of groups/variablesWhat is the average number of forum posts for each student?The answer to this question describes behavior in the forum.
RelationalAsks about relationships among variablesAre men more likely than women to post on forums?The answer to this question relates the gender of students to behavior in the forum.
CausalAsks about cause and effect relationships among variablesHow does the number of forum posts by the instructor affect the average number of posts by students?The answer to this question establishes cause and effect between instructor behavior and student behavior.
Table summarizing the types of research questions.

To view more posts about research design, see a list of topics on the Research Design: Series Introduction.

8 thoughts on “Research Design: Research Questions

  1. Pingback: Research Design: Series Introduction | Lauren Margulieux

  2. Thanks Lauren Margulieux for creating this content. I really enjoyed reading. My concern is about the question “What are the differences in learning outcomes between providing grades throughout the semester or providing only a final grade?” Why is it unethical?

    • I imagine the ethical concern about that question is related to giving students feedback about their progress throughout the semester. If a grade is provided only at the end, it’s plausible that a student wouldn’t realize that they were not understanding/performing well until the semester is over. Then they wouldn’t have any chance to change their behavior in order to understand/perform better.

  3. Pingback: Research Design: Non-Experimental and Experimental Designs | Lauren Margulieux

  4. Pingback: Research Design: Descriptive Statistics | Lauren Margulieux

  5. Pingback: Research Design: Inferential Statistics for Relational Questions | Lauren Margulieux

  6. Pingback: Research Design: Inferential Statistics for Causal Questions | Lauren Margulieux

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s