Crafting a Strong Hypothesis

Six points to consider when refining this important statement

We learn from the writings of Charles Darwin [1] how formal hypothesizing was central to his thinking and reasoning as he developed his ideas on natural selection. Today that approach is still as fundamental as ever (although there certainly is a place for research driven by objectives instead).

As pointed out by the NIH/NIAID in their guidance for submitting an application, the inclusion of a hypothesis is important because:

“..that’s what reviewers expect and what anchors your different Specific Aims to a common theme, not just a common field of research. Following a central hypothesis also keeps you focused with both writing the proposal and actually doing the research if the grant is funded.”

Despite its importance, the hypothesis can sometimes become an afterthought in the chaos of nailing down every other detail of a funding application or manuscript [2]. So as your work progresses, it’s worthwhile to reevaluate the strength of your hypotheses, and bring it back to the forefront of your projects.

In a 2022 review article on the topic of quantitative and qualitative research questions and hypotheses, Barroga and Matanguihan compiled a list of six characteristics of a well-developed hypothesis [3]. I’ve illustrated these below.

And to elaborate on these six points:

1. Empirically testable: A hypothesis should be stated in such a way that an experimental process can either support or refute the statement.
2. Backed by preliminary evidence: In the traditional scientific method, observations give way to hypotheses. Thus the central hypothesis of a research project should always be backed by preliminary experiments or observations gathered from previous work.
3. Testable by ethical research: This point adds a boundary to point #1, in that experiments, and thus hypotheses, are limited by what can be tested in an ethical manner. “First, do no harm,” etc.
4. Based on original ideas: New observations and ideas are the foundation to a novel hypothesis. If you can demonstrate that you are probably the first person to hypothesize something, it will distinguish and strengthen your project. In a world where everything claims to be “novel,” it is refreshing to see a project whose novelty speaks for itself.
5. Has evidence-based logical reasoning: This point overlaps with points #1 and #3. The traditional “if X…, then Y…” template of a hypothesis lends itself well to this point. Regardless of whether you use some version of that template, choose your words carefully so that they follow a logical path.
6. Can be predicted: Building on point #2, your audience should be presented with enough preliminary evidence or theoretical background that they would probably infer the same hypothesis before you present it to them. As Goldstein, et al., stated [4] in regards to drafting a Specific Aims page for NIH applications, “A well-written second paragraph should produce a narrative so clear that the reviewers can predict your hypothesis before they even read it.”

Diving a little deeper on these points, Barroga and Matanguihan add the following [3]:

“Good hypotheses can infer ethical and positive implications, indicating the presence of a relationship or effect relevant to the research theme. These are initially developed from a general theory and branch into specific hypotheses by deductive reasoning. In the absence of a theory to base the hypotheses, inductive reasoning based on specific observations or findings form more general hypotheses.”

Another consideration is the scope of a hypothesis, which is to say how broad (general) or narrow (specific) it should be. This question was explored in deep philosophical detail by Thompson and Skau in a 2023 paper [5]. They suggest a model in which a hypothesis has three dimensions that define its scope: the variables, the relationships between those variables, and the analytical process used to evaluate them (i.e., ‘pipeline’). Most importantly though, they posit that broad or narrow hypotheses can both be valid, depending on the context.

A good example of where the scope of a hypotheses comes into play might be in a typical NIH Specific Aims page. This will usually include a central hypothesis that the specific aims will test. This central hypothesis should be broad, general, and overarching, such that any additional hypotheses stated in the subsequent aims will be related to it in a distinct and more specific way.

Hopefully these considerations can be helpful as you continue developing and refining the hypotheses that guide your research. Doing this may not only help demonstrate the value of your research to reviewers, but also be useful for you in identifying new and interesting research questions.

References:

1. National Academy of Sciences (US). In the Light of Evolution: Volume III: Two Centuries of Darwin. Avise JC, Ayala FJ, editors. Washington (DC): National Academies Press (US); 2009. PMID: 25032348.
2. Kerr NL. HARKing: hypothesizing after the results are known. Pers Soc Psychol Rev. 1998;2(3):196–217. doi: 10.1207/s15327957pspr0203_4. PMID: 15647155.
3. Barroga E, Matanguihan GJ. A Practical Guide to Writing Quantitative and Qualitative Research Questions and Hypotheses in Scholarly Articles. J Korean Med Sci. 2022 Apr 25;37(16):e121. doi: 10.3346/jkms.2022.37.e121. PMID: 35470596; PMCID: PMC9039193.
4. Goldstein AM, Balaji S, Ghaferi AA, Gosain A, Maggard-Gibbons M, Zuckerbraun B, Keswani SG. An algorithmic approach to an impactful specific aims page. Surgery. 2021 Apr;169(4):816–820. PMID: 32709487
5. Thompson WH, Skau S. On the scope of scientific hypotheses. R Soc Open Sci. 2023 Aug 30;10(8):230607. doi: 10.1098/rsos.230607. PMID: 37650069; PMCID: PMC10465209.

Robert Lawrence is a science editor at Baylor College of Medicine. You can find his published work at www.robertlawrencephd.com