By Rajiv S. Jhangiani, I-Chant A. Chiang, Carrie Cuttler and Dana C. Leighton, adapted by Marc Chao and Muhamad Alif Bin Ibrahim

Formulating an empirically testable research question is a crucial step in the scientific process. It transforms a general research idea into a focused inquiry that can be investigated systematically. At its core, an empirically testable question must involve observable and measurable variables, either focusing on a single variable or exploring the relationship between multiple variables. While this process may seem daunting at first, as though experienced researchers pluck compelling questions out of thin air, it is, in reality, the result of strategic thinking, persistence, and familiarity with the research literature.

One effective approach to generating research questions is to examine the discussion sections of recent academic articles. The discussion section is where researchers interpret their results, relate them to past studies, and suggest directions for future research. These suggestions often highlight unanswered questions, methodological limitations, or intriguing findings that warrant further exploration. Because these future directions have already been flagged by experienced researchers as meaningful and important, they offer fertile ground for developing your own research questions.

Beyond reviewing existing studies, you can also generate research questions by starting with a specific behaviour or psychological characteristic and framing it as a variable. For example, you might ask: How many words do people speak in a day? How accurate are people’s memories of traumatic events? What percentage of adults experience chronic anxiety? If these questions have not yet been thoroughly explored, as you will discover during your literature review, they might represent valuable research opportunities.

If a variable has already been studied extensively, the next step is to consider relationships between variables. For instance, you might ask what factors cause a particular behaviour or psychological characteristic, what consequences it might have, or how it varies across different people or situations. If you are interested in talkativeness, you might ask whether family size influences how much people talk, or whether same-sex social groups foster more conversation than mixed-sex groups. Each potential relationship represents a unique research question that could contribute to the broader scientific understanding of your topic.

However, encountering a question that has already been answered by previous research does not mean you should abandon it. Instead, consider refining the question to offer a fresh perspective or address a gap in the literature. Are there alternative ways to define or measure the variables in question? Are there specific populations where the relationship might be stronger or weaker? Could situational factors influence the outcome in meaningful ways? For example, while previous research suggests men and women speak about the same number of words per day, you might refine the question by focusing on whether this finding holds true across different age groups or cultural contexts. Through this process, even well-explored topics can yield new avenues for investigation.

Evaluating Research Questions

Generating a list of potential research questions is only the beginning. Researchers must carefully evaluate each question to determine which are worth pursuing. Two essential criteria guide this evaluation: interestingness and feasibility.

Interestingness

A research question’s interestingness is not about whether it fascinates you personally, but whether it holds broader relevance for the scientific community. Several factors determine this.

First, a research question is interesting if its answer is genuinely in doubt. Questions that have already been conclusively answered through prior research are no longer compelling subjects for new investigation. However, if reasonable arguments can be made for multiple potential answers, the question becomes much more engaging. For example, the question of whether women are more talkative than men is intriguing because plausible arguments exist on both sides: stereotypes suggest women are more talkative, but evidence shows little difference in verbal abilities between genders.

Second, a question is interesting if it fills a gap in the existing literature. Even if a question has not been answered empirically, it must feel like a natural and meaningful extension of what is already known. For example, asking whether taking notes by hand improves academic performance naturally follows from research showing the cognitive benefits of deeper information processing.

Finally, a research question gains significance if it has practical implications. Questions that address real-world problems or inform practical decision-making are often considered more valuable. For instance, exploring whether cell phone use impairs driving performance carries meaningful consequences for public safety and policy-making.

Feasibility

A research question might be theoretically fascinating, but if it cannot realistically be answered given your resources, expertise, or timeline, it is not worth pursuing. Feasibility depends on several factors, including time, funding, access to equipment, technical skills, and availability of participants.

For instance, a large-scale longitudinal study tracking participants over decades would require significant time and funding, resources typically unavailable to a single student researcher. Similarly, a neuroimaging study involving advanced brain-scanning technologies may not be feasible without access to specialised labs and training.

However, feasibility does not mean compromising on the quality of your study. Many impactful studies are relatively simple and resource-friendly, relying on university student samples or straightforward observational tasks. Even small-scale studies can yield meaningful contributions if they are well-designed and methodologically sound.

When designing your study, it is often wise to borrow methods from existing research. If previous studies have successfully manipulated participants’ moods by offering compliments, for example, adopting this approach is both practical and methodologically consistent. Not only does this increase the likelihood of success, but it also ensures your findings are easier to compare with existing literature.

Theories and Hypotheses

Understanding the difference between a theory and a hypothesis is essential for conducting meaningful scientific research. While these two terms are often used interchangeably in everyday conversation, they have distinct meanings in the realm of science. A theory is a coherent and systematic explanation of one or more phenomena, built upon established evidence and reasoning. It serves as a framework for understanding and predicting outcomes. Theories often introduce abstract concepts, relationships, and processes that go beyond the observable data.

For example, Zajonc’s theory of social facilitation and social inhibition (1965) suggests that being observed by others during a task creates a state of physiological arousal. This arousal, in turn, enhances the performance of well-practised tasks (social facilitation) but impairs performance on unfamiliar or complex tasks (social inhibition). The theory introduces terms like arousal and dominant response, which are not directly observable but serve as essential constructs for explaining observed behaviour patterns. Such theoretical constructs provide a foundation for generating specific hypotheses and guiding further research.

It is important to note that in science, the term theory does not imply uncertainty or guesswork, as it often does in everyday language. A scientific theory can be extensively tested, well-supported, and widely accepted by the scientific community. For instance, the theory of evolution by natural selection and the germ theory of disease are both referred to as theories, not because they are speculative, but because they provide comprehensive explanations for large sets of observed phenomena. These theories are supported by vast amounts of empirical evidence and continue to guide scientific discovery.

In contrast, a hypothesis is a specific, testable prediction about what should be observed if a theory is accurate. Hypotheses are narrower in scope and focus on particular aspects of a theory or phenomena. They are formulated based on existing evidence, logical reasoning, or theoretical frameworks and are often stated in ways that allow them to be tested empirically. For example, based on Zajonc’s drive theory, one might hypothesise: If drive theory is correct, then cockroaches should run faster through a straight runway but slower through a branching runway when other cockroaches are present.

However, not all hypotheses are derived from existing theories. In some cases, researchers generate atheoretical hypotheses, which arise from observations or preliminary data without being directly tied to an overarching theory. For example, if researchers notice an unexpected behavioural pattern during preliminary observations, they might develop a hypothesis to investigate that pattern further. Over time, a broader theory might emerge from a collection of related hypotheses and findings.

Hypotheses often take the form of if-then statements, establishing a clear relationship between variables. For example, if expressive writing helps people habituate to negative emotions, then writing about traumatic experiences should reduce emotional distress more effectively than writing about positive experiences. Even when stated as declarative sentences, hypotheses can always be rephrased as research questions, such as “Does expressive writing about traumatic experiences reduce emotional distress more than writing about positive experiences?”

Deriving Hypotheses from Theories

The process of generating hypotheses from theories typically begins with identifying a research question. Researchers can then ask whether any existing theory provides a potential answer to that question. For instance, if a researcher wonders whether writing about positive life events has the same psychological benefits as writing about traumatic events, they might turn to habituation theory. According to habituation theory, emotional benefits arise from repeated exposure to negative thoughts and feelings, which reduces their emotional impact over time. If this theory is correct, writing about positive experiences should not yield the same benefits as writing about traumatic experiences because positive events do not evoke distress that requires habituation.

Another way to derive hypotheses from theories is to focus on specific components or mechanisms within the theory that have not yet been directly observed or tested. For example, a researcher could examine whether emotional habituation happens gradually across multiple expressive writing sessions by measuring participants’ distress levels after each session.

Among the most valuable hypotheses are those that can distinguish between competing theories. For example, Norbert Schwarz and his colleagues (1991) investigated two competing theories about how people judge their assertiveness. One theory proposed that people base their self-judgements on the number of relevant examples they can recall, while the other theory suggested that judgements are based on how easily those examples come to mind. To test these theories, participants were asked to recall either six (easy) or twelve (difficult) examples of their assertive behaviour and then rate their overall assertiveness. The first theory predicted that recalling more examples would lead to higher assertiveness ratings, while the second theory predicted that ease of recall would play a more significant role. The results supported the ease-of-retrieval theory, demonstrating the value of crafting hypotheses that pit one theory against another.

Theory Testing and the Hypothetico-Deductive Method

The process of testing theories follows a method known as the hypothetico-deductive approach. Researchers start with an existing theory or construct one based on observed phenomena. From this theory, they derive specific hypotheses, which are predictions about what should occur under certain conditions if the theory is accurate. They then design and conduct empirical studies to test these hypotheses. Based on the results, the theory is either supported, refined, or revised. This cyclical process, as shown in Figure 6.4.1, is essential for advancing scientific understanding, as each iteration builds upon the findings of previous research.

A classic example of this approach comes from Zajonc’s research on social facilitation and inhibition. After developing drive theory, Zajonc hypothesised that cockroaches would perform better on simple tasks and worse on complex tasks when observed by others. His experiments confirmed these predictions, providing strong empirical support for his theory. This iterative process demonstrates how theory-driven research contributes to scientific progress by refining and expanding theoretical frameworks.

Incorporating Theory into Your Research

Incorporating theory into your research enhances its significance and clarity. There are two primary ways researchers typically use theories in their work. The first approach involves conducting a study to answer a research question and then using one or more theories to interpret the results. This approach is particularly useful for applied research or when existing theories do not directly address the question at hand. The second approach involves deriving a hypothesis from an existing theory, testing that hypothesis through an empirical study, and then evaluating or refining the theory based on the results.

Using established theories not only strengthens the foundation of your research but also situates your work within the broader scientific dialogue. Psychological theories are the result of decades of research and represent collective knowledge about human behaviour and mental processes. By aligning your research with these theoretical frameworks, you ensure that your findings contribute meaningfully to the scientific community.

Characteristics of a Good Hypothesis

A strong hypothesis possesses three key characteristics: testability, logical reasoning, and positivity.

First, a hypothesis must be testable and falsifiable. This means it must be possible to gather empirical evidence that could disprove the hypothesis if it is incorrect. If a hypothesis cannot be tested or proven false, it falls outside the realm of scientific inquiry.

Second, a hypothesis must be logical. It should be informed by existing theories, observations, or empirical data and should follow a clear line of reasoning. Hypotheses are not random guesses; they emerge from a structured thought process that connects prior knowledge to new questions.

Finally, a hypothesis should be positive. It should make a statement about the existence of a relationship or effect rather than the absence of one. Scientists begin with the assumption that no effect exists (the null hypothesis) and then look for evidence to reject this assumption in favour of an alternative hypothesis.

By crafting hypotheses that are testable, logical, and positive, researchers create clear, focused questions that can be systematically investigated, ultimately advancing our understanding of complex phenomena.

References

Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer-Schatka, H., & Simons, A. (1991). Ease of retrieval as information: Another look at the availability heuristic. Journal of Personality and Social Psychology, 61(2), 195–202. https://doi.org/10.1037/0022-3514.61.2.195