3 Chapter 3: Research Questions

Learning Objectives:

• Understand the iterative nature of research design and the relationship between research questions, hypotheses, and study design.
• Identify sources of inspiration for research questions and evaluate the feasibility and impact of research ideas.
• Differentiate between research questions and hypotheses and understand the importance of variables in their formulation.
• Learn how to ground research direction in scientific literature and use it to refine research questions and hypotheses.
• Develop skills in formulating focused, researchable, and relevant research questions and hypotheses.

Key Terms:

• Variable: A concept or characteristic that can be measured, quantified, or otherwise observed. Quantitative measurement of variables involves numerical representation of a variable; qualitative measurement are non-number categories of a variable (e.g., text, images, videos). Dependent variables are the outcomes of focus in the research study (also called response variables). Independent variables are the factors that can influence the dependent variable. Demographic variables are population characteristics of research participants.
• Research question: An inquiry into an unexplored or contested area of science, and it guides a research project and is formed with variables. Exploratory research questions are open-ended inquiries that aim to investigate a broad topic without a predetermined hypothesis; confirmatory research questions are specific inquiries designed to test a predefined hypothesis or theory.
• Hypothesis: A specific, testable prediction derived from a research question and formed with variables. A falsifiable hypothesis must be structured in a way that allows it to be proven false through empirical observation or experimentation.
• Scientific and peer-reviewed literature: Scientific literature are written works that report on original scientific research, theories, or reviews of existing research in the field of science, serving as a foundation for new studies; peer-reviewed literature are studies that have passed scientific merit reviews of experts in the field.

Introduction

In this chapter, we will explore how to narrow down broad research interests into precise questions, define variables, and utilize literature to inform the direction of our research. By the end of this chapter, you will be equipped with the tools to craft well-defined research questions and hypotheses that are grounded in scientific literature and poised for empirical investigation.

Iterative Nature of Research Design

Research in the health sciences is a much more of a spiral than a straight line. This section underscores the iterative nature of research design in which the researcher simultaneously develops research questions (this chapter) and research designs (next chapter). While it can be frustrating to repeatedly revise the research question based on the changes to the research design, and vice-versa, this process ensures that the study design can answer the research question.

A research question is a guiding inquiry concerning what the study seeks to explore or uncover. It serves as the foundation for the investigation. A hypothesis, meanwhile, is a specific, testable prediction derived from a research question. It proposes a potential outcome based on the theoretical framework and existing literature. Further below we will clarify these definitions of research questions and hypothesis.

Figure 3: Grounding Questions in Scientific Literature

The formulation of a research question or hypothesis is deeply intertwined with the selection of a study design. As hypotheses take shape from the review of literature and theoretical considerations, they inform the choice of study design that is most appropriate for testing them. Conversely, the practicalities and constraints of potential study designs can lead to further honing of the question. This cyclical relationship ensures that the research question, hypothesis, and study design are cohesively aligned, allowing for the most effective and efficient path to answers.

In the next sections, we will explore how to narrow down broad research interests into precise questions, define independent and dependent variables, and utilize literature to inform the direction of our research.

Starting with a Broad Research Topic

Drawing Inspiration From… Anywhere

Health research often begins with broad, exploratory questions that seek to understand complex real-world observations. Researchers draw inspiration for these questions from a range of sources, sometimes coming from our personal or professional lives. These sources can include personal experiences and perspectives, academic literature, news outlets, and social media. For example, one might notice an increasing discussion around mental health and its relation to dietary habits on various platforms, inspiring research questions on links between nutrition and mental health.

Students searching for research topics can consider these possible sources of ideas:

• Your personal experiences or questions about health.
• Noticing patterns or health trends in your community or environment.
• Bringing your health questions/thoughts to conversations with artificial intelligence for further insights and brainstorming, such as Microsoft Copilot and Google Gemini.
• Utilizing hashtag searches on social media to gauge public interest or concern about specific health issues.
• Assessing worldwide health reports or WHO announcements for emerging health concerns.
• Listening to podcasts on health-focused discussions often highlight contemporary issues and research gaps.

Evaluating Ideas for a Research Project

Once researchers have broad ideas for areas of investigation the next step is to evaluate the merit of the idea for a research project. Determining the feasibility and worthiness of a research idea involves several criteria:

• Relevance: Is the topic of widespread interest within both the scientific community and society at large?
• Literature Base: Does a substantial amount of research already exist on this topic, indicating its recognition and importance?
• Timeliness: Is the topic timely, addressing current issues or gaps in scientific discourse?
• Feasibility and Impact: Consider if the research is feasible given available resources and if it has the potential to make a meaningful impact.

For students going through this exercise, consider the above list as non-exhaustive and there are likely other context-based factors to consider. The evaluation stage is further clarified by developing measurable research questions and/or hypotheses.

Exploratory vs Confirmatory Research

Research questions and hypotheses are further characterized by being exploratory or confirmatory in nature (Schwab and Held 2020).

• Exploratory Research: Hypothesis-free research, also known as exploratory research, is characterized by the absence of predefined hypotheses. This approach is particularly valuable in the early stages of research when the relationships between variables are not well understood. It allows researchers to explore large datasets, identify patterns, and generate new hypotheses based on the observed data. Unlike hypothesis-driven research, which tests specific predictions, hypothesis-free research is open-ended and driven by the data itself. For example, in genomics research, scientists often employ hypothesis-free approaches to analyze vast amounts of genetic data, looking for associations that might indicate a link between specific genes and diseases. Similarly, in qualitative research, researchers may conduct interviews or focus groups without a predetermined hypothesis, allowing themes and patterns to emerge organically from the data.
• Confirmatory Research: In contrast, confirmatory research is conducted to test specific hypotheses and theories. It is more structured and uses quantitative methods to collect and analyze data. Confirmatory research seeks to confirm or disprove existing theories, providing empirical evidence to support or refute them. This type of research is essential for validating scientific theories and advancing our understanding of health outcomes. For example, to confirm whether a new drug (Drug A) is more effective than a standard treatment (Drug B) in reducing symptoms of depression.

By understanding the differences between exploratory and confirmatory research, researchers can choose the appropriate approach for their study, depending on the stage of knowledge in their field and the specific research questions they aim to answer.

Research Questions with Variables

Once researchers have identified a broad research area and either an exploratory or confirmatory direction, the next step is to develop specific research questions and/or hypotheses in the area of investigation. Above, we identified a scientific research question as a fundamental inquiry that guides a research project, as the research question shapes the entire research process. Let’s specify that a scientific research question must include variables. Variables are essential for conducting empirical research, as they allow researchers to systematically investigate relationships, effects, and causes within a study.

A variable is a concept or characteristic that can be measured, quantified, or otherwise observed, and they are elements or factors within the research that can change or vary. All variables in a research question or hypothesis should be tied to clearly defined measures. Table 2 presents some examples of variables.

Table 2: Examples of Variables
Variable	Measurement
Student’s grades	GPA, letter grades, percentage scores
Cardiovascular health	Blood pressure, cholesterol levels, heart rate variation
Attitudes towards abortion	Survey responses on a Likert scale
Happiness	Self-reported happiness scale, psychological assessments

Quantitative vs Qualitative Measurement

Measurement of variables is done quantitatively or qualitatively. Quantitative measurement involves the process of quantifying variables using numerical data, which allows for statistical analysis. Quantitative measurements are used in research to generalize from a sample to a larger population, such that researchers can estimate the prevalence of certain health conditions or factors in a population.

Qualitative measurement focuses on capturing non-numerical data, such as text, images, audio, and video, which often related to experiences, beliefs, perceptions, or behaviors, to gain an in-depth understanding of complex phenomena, such as through interviews. This approach places emphasis on context, meaning, and depth of information, and less emphasis on generalizability to larger populations.

The differences in data collection and analysis between these two approaches are significant. Quantitative measurement typically involves statistical analysis of numerical data, allowing researchers to identify scaled/exact patterns and relationships among variables. In contrast, qualitative measurement involves methods such as thematic analysis and content analysis to interpret patterns in non-numerical data, providing context and detailed insights into the subject matter.

Figure 4: Strengths & Weaknesses of Quantitative vs Qualitative Measurement. Image Credit: Warren, K. (2020). Qualitative Data Analysis Methods 101. Gradcoach.com, accessed March 20, 2024.

Strengths & Limitations

Generally, quantitative methods, often associated with larger sample sizes, excel in providing breadth of information. Sample size refers to the number of participants in a study, which serves to represent the research population of a study, which the group of humans that a study seeks to gain knowledge. Quantitative measures offer a wide view that can be generalized to a larger population but may lack depth in individual experiences. Conversely, qualitative methods excel in depth, offering rich, detailed insights into fewer individuals’ experiences. While they provide a profound understanding of phenomena, they may not be as easily generalized across a broader population due to smaller sample sizes. Each method’s strength simultaneously serves as its weakness; the quantitative approach’s breadth may overlook nuances, while the qualitative approach’s depth may limit its applicability.

The integration of quantitative and qualitative measurements, known as mixed-methods research, combines the strengths of both approaches to provide a more comprehensive understanding of a research problem. This approach allows researchers to explore the breadth and depth of a topic, capturing both the numerical and experiential aspects of the variables under study.

When choosing between quantitative and qualitative measurement, researchers must consider several factors. Alignment with research objectives is crucial, as some questions are best answered through numerical data, while others require an exploration of subjective experiences. The feasibility and resources available for data collection and analysis also play a significant role in determining the appropriate approach. Additionally, the suitability of the measurement method for the research question is essential, as it impacts the validity and reliability of the findings (further discussed in Chapter 7 on instrumentation).

Independent vs Dependent Variables

Let’s consider four important categories of variables that are necessary to constructing research questions and hypotheses: independent, dependent, and demographic.

• Independent Variables: These are the factors that you, as a researcher, manipulate or vary to observe their effect on other variables. For instance, in studying the impact of dietary habits on mental health, the type of diet (e.g., high in processed foods vs. high in fruits and vegetables) would be an independent variable.
• Dependent Variables: These are the outcomes or effects that are measured to see how they change in response to alterations in the independent variable. Continuing with the example, measures of mental health wellness would be the dependent variable, as they are expected to vary based on dietary intake.
• Demographic Variables: These refer to characteristics of the study population that might influence the research outcome, such as age, gender, socioeconomic status, etc. These variables are crucial for understanding and interpreting the nuances in research findings.

All research questions have dependent and demographic variables, and most research questions have independent and control variables. Below is a step-by-step guide to help students to refine a broad research area into specific, measurable research questions.

• Identify Your Independent Variables: Sometimes researchers start by pinpointing the factors they anticipate will influence outcomes in the study. For example, if you’re interested in the effects of diet on mental health, your independent variable could be the type of diet (e.g., high in processed foods vs. high in fruits and vegetables).
• Determine Your Dependent Variables: Sometimes researchers start by defining a dependent variable, or if they start with independent variables, then the dependent is immediately defined afterwards. Identify the variable you will study that is being effected of your independent variables (or that you believe is being effected by the independent variables). In the diet and mental health example, the dependent variable could be indicators of mental health wellness, such as anxiety levels or mood.
• Consider Demographic Variables: Demographics are population characteristics, such as age, sex, race, income, and much more. Demographic variables are often necessary in research questions to specify research populations, which can significantly influence your study’s outcomes.
• Formulating Your Research Question: With a clear understanding of your variables, you can now formulate a focused research question. For instance, “How does a diet high in fruits and vegetables compared to a diet high in processed foods affect anxiety levels among adults aged 20-30?”

Table 3 presents examples using ABCD of narrowing down broad research areas into research questions with variables. Lastly, below are general best practices for crafting well-constructed research question:

• Focused: A good research question is specific and clearly defines the scope of the study. It avoids being too broad or vague.
• Researchable: The question should be answerable through empirical evidence, data collection, observation, or experimentation. It should be feasible to investigate using research methods.
• Relevant: The research question should address a timely social or scholarly issue. It should contribute to existing knowledge or provide insights into a specific problem.
• Clear Criteria: The question should use clearly defined terms and concepts. Avoid normative questions (e.g., “What should be done?”) unless you’re exploring possible solutions.
• Originality: While some research questions explore well-established topics, consider identifying underexplored aspects or gaps in existing knowledge.

Hypotheses with Variables

Earlier we identified hypotheses as testable predictions about outcomes of research questions. Incorporating variables into a hypothesis adds specificity by clearly identifying what will be measured (dependent variable) and what will influence those measurements (independent variable). For instance, “Increased physical activity (independent variable) will improve cognitive function (dependent variable) in adolescents.” By setting clear expectations for study results, hypotheses facilitate the application of statistical analysis to confirm or refute these predictions. Table 3 contrasts measurable research questions and hypotheses.

When transitioning from research questions to hypotheses, it is important to make a clear distinction between the two. Hypotheses are derived from research questions and provide a focused framework for investigation. They should be clearly stated, with explicit mention of the independent and dependent variables, and should be formulated in a way that allows for empirical testing.

Falsifiable Hypotheses: The Pillar of Empirical Research

A key characteristic of a robust hypothesis is its falsifiability. A falsifiable hypothesis is one that can be proven false through empirical observation or experimentation. This concept, introduced by philosopher Karl Popper, is a critical criterion for distinguishing scientific hypotheses from non-scientific ones.

To write a falsifiable hypothesis:

• Clearly Define Variables: Ensure that the variables are clearly defined and measurable.
• Specify Predictions: Make specific predictions about the relationship between variables that can be unequivocally supported or refuted by empirical data.
• Avoid Vague Language: Use clear and concise language to eliminate ambiguity.
• Consider Alternative Outcomes: Think about possible alternative outcomes and how they would challenge the hypothesis.
• Use Conditional Statements: Express hypotheses as conditional statements (e.g., “If X, then Y”) to highlight the cause-and-effect relationship being tested.

For example, a non-falsifiable hypothesis might state, “Exercise improves well-being.” This is vague and difficult to test. A falsifiable version could be, “Engaging in 30 minutes of moderate aerobic exercise three times a week for eight weeks will lead to a statistically significant decrease in self-reported stress levels among adults aged 18-65.” This hypothesis is specific, measurable, and can be proven false if the exercise regimen does not lead to the predicted decrease in stress levels.

By ensuring that hypotheses are both specific to the variables involved and falsifiable, researchers can create a solid foundation for empirical investigation that contributes to the accumulation of reliable and valid scientific knowledge.

Null Hypothesis

The null hypothesis is a specific type of falsifiable hypothesis that is common throughout research in the sciences. The null hypothesis asserts that there is no effect, no difference, or no relationship between the variables under investigation. Essentially, it serves as a default position that the research hypothesis aims to challenge. In hypothesis testing, statistical methods are used to determine whether the observed data are consistent with the null hypothesis or if there is sufficient evidence to support the research hypothesis, commonly referred to as an alternative hypothesis. Here’s a breakdown of how they work together:

• Null Hypothesis (H0): This hypothesis posits that there is no significant difference or relationship between groups or variables being studied. It is the hypothesis that the study aims to test directly.
• Alternative Hypothesis (H1 or Ha): This hypothesis is the counterpart to the null hypothesis. It states that there is a significant difference or relationship between the groups or variables. The alternative hypothesis is what the researcher really wants to prove or support.

In a typical hypothesis-testing scenario, the null hypothesis is tested with the intention of rejecting it. If the null hypothesis is rejected after statistical analysis, it suggests that there is enough evidence to support the alternative hypothesis. The alternative hypothesis is considered only when the null hypothesis is found to be unlikely given the data.

For example, in a clinical trial to test a new medication:

• Null Hypothesis (H0): The medication has no effect on the disease’s symptoms compared to a placebo.
• Alternative Hypothesis (H1): The medication has a positive effect on reducing the disease’s symptoms compared to a placebo.

If the results show statistically significant improvement in symptoms with the medication compared to the placebo, the null hypothesis is rejected, and the alternative hypothesis is supported.

The PICO Format for Hypotheses

In health research, particularly in clinical studies, the PICO format is a widely used tool for formulating research questions and hypotheses, originating in Evidenced-Based Medicine practices (Straus 2011). PICO stands for Patient population, Intervention, Comparison, and Outcome. This format helps researchers to structure their hypotheses in a way that is specific, focused, and relevant to clinical practice.

• Patient population (P): This element specifies the group of patients or individuals who are the subject of the research. It defines the characteristics of the population, such as age, gender, health condition, or any other criteria relevant to the study.
• Intervention (I): This component describes the treatment, procedure, or exposure that is being investigated. It could be a medical intervention, a behavioral strategy, a dietary change, or any other factor that is being applied or examined.
• Comparison (C): The comparison refers to the alternative to the intervention, against which the intervention’s effects are measured. This could be a placebo, standard care, a different treatment, or no treatment at all, depending on the research question.
• Outcome (O): The outcome is the effect or result that is being measured in the study. It could be a clinical outcome, such as improvement in symptoms, a change in biomarkers, or any other measurable consequence of the intervention.

For example, a hypothesis structured using the PICO format might be: “In adults aged 40-60 with type 2 diabetes (P), a low-carbohydrate diet (I) compared to a low-fat diet (C) leads to greater improvement in glycemic control (O) over a six-month period.” The clarity of the PICO format helps in designing the study, selecting appropriate measurement tools, and analyzing the results.

Table 3: Relationships Among Broad Research Areas and Example Research Questions and Hypotheses in the ABCD Study
Broad Research Areas	Specific Research Questions	Falsifiable Research Hypotheses
Characterize individual developmental trajectories (e.g., brain, cognitive, emotional, academic) and the factors that can affect them.	How does varying screen time before bed affect cognitive development and emotional well-being in adolescents?	Adolescents with increased screen time before bed will exhibit lower cognitive development scores and poorer emotional well-being compared to adolescents with limited screen time.
	How does the frequency of family meals relate to social development and academic performance in adolescents?	Adolescents who frequently participate in family meals will show enhanced social development and higher academic performance compared to their peers with fewer family meals.
Develop national standards of normal brain development in youth.	What are the correlations between physical activity levels and brain development metrics in youth aged 10-15?	Youth aged 10-15 who engage in regular physical activity will show higher brain development metrics than those who are less active.
	How does adherence to national nutritional guidelines correlate with neurodevelopmental milestones in youth aged 10-15?	Youth aged 10-15 who adhere to national nutritional guidelines will achieve neurodevelopmental milestones more effectively than those who do not follow these guidelines.
Disentangle genetic vs. environmental factors on development.	How do genetic predispositions to ADHD and the home learning environment interact to affect cognitive development in youth?	The cognitive development in youth with genetic predispositions to ADHD is significantly influenced by the quality of their home learning environment, with a positive environment mitigating some of the potential developmental challenges.
Examine effects of physical activity, sleep, screen time, and other activities on brain developmental outcomes.	How does screen time affect social and brain development?	Increased screen time is negatively correlated with social development and alters brain structure in pre-teens.
	How do sleep patterns affect academic achievement?	Adolescents with regular, sufficient sleep patterns will exhibit higher academic achievement than those with irregular or insufficient sleep patterns.
Study the onset and progression of mental health disorders and their influencing factors.	What is the impact of early-life stressors on the risk of developing depressive disorders during adolescence?	Adolescents exposed to high levels of early-life stressors are at greater risk of developing depressive disorders.
	What role do extracurricular activities play in mitigating the onset of anxiety disorders in high school students?	Participation in extracurricular activities reduces the risk of developing anxiety disorders in high school students compared to students who do not participate in such activities.
Determine how exposure to various levels and patterns of alcohol, nicotine, cannabis, caffeine, and other substances affects developmental outcomes and vice versa.	What are the long-term effects of ADHD medications on academics and health?	Long-term use of ADHD medications in children and adolescents is associated with improved academic performance but may have mixed effects on physical health outcomes.
	How does tobacco or alcohol use affect learning and health?	Adolescents who use tobacco or alcohol will show lower learning outcomes and adverse health effects compared to their non-using peers.

Grounding the Research Direction in Scientific Literature

The previous section outlined the mechanics of writing research questions and hypotheses, and this section defines the context of the process. Specifically, researchers must ground their research questions and hypotheses in the scientific and/or peer-reviewed literature. This process is also called a literature review, which is discussed in depth in Chapter 6, and it serves it justify the proposed research questions and hypotheses.

Scientific literature refers to written works that report on original scientific research, theories, or reviews of existing research in the field of science. Scientific literature can include articles in scientific journals, conference papers, and books that present empirical findings, methodologies, and theoretical contributions to the field. Peer-reviewed literature include scientific literature that have undergone a formal evaluation process by experts or peers in the same field before being published. The peer review process is a quality control mechanism that ensures the research is sound, the methodology is appropriate, and the conclusions are valid. Peer-reviewed literature is considered a gold standard in academic publishing and is often required for academic and professional credibility.

Using Literature to Shaping Initial Ideas

At this point, students will have a general intuition for writing research questions and hypotheses, but how do we judge the significance or impact of a research question? Researchers do this by situating research questions in the context of the scientific literature. An easy way to find scientific literature is by using Google Scholar search, and Chapter 6 provides a wider list of resources and processes in literature search and review.

Researchers use the scientific literature to further refine a research question or hypothesis. Students at this stage can take this step by reading through the articles, pay particular attention to sections of an article containing research questions, research purposes, and hypotheses, and discussion and limitations. Some starter considerations for students include:

• Could existing research questions or hypotheses be modified or updated to a create a new research question or hypothesis?
• Could existing research (and research questions and hypotheses) benefit from being tested in different contexts, with different measures, or with demographically different participants?
• Do existing studies outline their limitations (typically found in discussion sections), and would you propose research questions to help overcome any limitations in previous research?
• Do existing studies propose directions for future research (typically found in discussion sections), and does this help inform your research question?

Beyond refining ideas, researchers also situate their research questions within conventions and theories of a discipline. Considerations here include:

• How has the research topic been studied over time, and to what extent is your research question consistent with the theoretical and methodological trends in the topic?
• What are the “landmark” or most frequently cited studies in the research topic or field, and how does your topic relate to them?
• Have certain theoretical or methodological approaches been more successful than others, and how does this relate to your proposal?
• What are the relevant theories and their applications in research topic, and will your study contribute to theoretical development (such as providing empirical tests of the theories)?

Hypotheses Formulation

After understanding the nuts and bolts of research questions and hypotheses, and why and how to use scientific literature to refine and contextualize them, students are ready to advance their hypotheses formation. The next set of considerations is to match the research questions and hypotheses to the study design, which is the topic of the next chapter. As such, generating research questions and hypotheses is an iterative process that is subject to much refinement and revision throughout the planning phases of research.

Summary

This chapter explored processes of formulating research questions and hypotheses, which serve as the foundation for any research project. We emphasized the iterative nature of research design, where research questions and hypotheses are refined in tandem with the development of the study design. The chapter provides guidance on finding inspiration for research questions, evaluating the feasibility of research ideas, and the importance of grounding research in scientific literature. It also highlights the significance of variables in shaping research questions and hypotheses and offers a step-by-step approach to formulating focused, researchable, and relevant inquiries.