Pragmatism Offers a Better Way to Do Research

The Problem with “The Method”

In 2017, a team of social psychologists published a paper showing that people who held a warm cup of coffee judged a stranger as having a warmer personality. The finding was elegant, intuitive, and almost certainly false. When other labs tried to replicate it, the effect vanished. The coffee study became a poster child for the replication crisis, and for good reason. But it also exposed a deeper problem that nobody wanted to name: the way we teach research methods is broken.

Most researchers learn methodology the way children learn to swim: by being thrown into the deep end. They memorize the rules of null hypothesis testing, the etiquette of qualitative coding, the proper way to write a survey. They learn that quantitative methods are for proving things and qualitative methods are for understanding things. They learn that you pick your method first, then ask your question. The method becomes the master, not the servant.

Alex Gillespie, Vlad Petre Glăveanu, and Constance de Saint Laurent have a better idea. In their 2024 book “Pragmatism and Methodology,” they argue that the entire edifice of research methodology needs to be rebuilt from the ground up (Gillespie et al., 2024). Their foundation is pragmatism, the American philosophical tradition that judges ideas by their consequences rather than their origins. And their target is nothing less than the way we think about what counts as good research.

Why Are We Still Fighting About Numbers vs. Stories?

The quantitative qualitative divide is one of the most persistent and destructive features of the social sciences. It is also, according to Gillespie and colleagues, entirely unnecessary. The authors trace this divide back to a philosophical mistake: the belief that methods come with built in ontologies, that using numbers commits you to a view of reality as mechanical and objective, and that using words commits you to a view of reality as subjective and constructed.

This is nonsense, they argue. A number is just a tool. So is a narrative. What matters is what you do with them.

The pragmatist approach starts with a simple question: What are you trying to accomplish? If you want to know whether a new teaching program improves test scores, you need numbers. If you want to know how students experience that program, you need stories. If you want to know both, you need both. The method follows the question, not the other way around.

Gillespie and colleagues call this “methodological pragmatism” (Gillespie et al., 2024). It is not a compromise between quantitative and qualitative approaches. It is a rejection of the entire framework that pits them against each other. The authors argue that the real divide is not between numbers and words but between useful and useless research. A study that answers a meaningful question is good, regardless of whether it uses a t test or a thematic analysis. A study that answers a trivial question is bad, regardless of its methodological sophistication.

What If Your Question Changed Midway?

Here is where pragmatism gets genuinely radical. Most research training assumes that the research question comes first, the method second, and the data third. This linear model is tidy. It is also, according to Gillespie and colleagues, almost never how real research works.

In practice, questions evolve. You start with one question, collect some data, and discover that the question was wrong. Or you find a surprising pattern that demands a new question. Or you realize that your original method cannot actually answer what you want to know. The linear model treats these moments as failures. Pragmatism treats them as opportunities.

The authors describe this as “abductive reasoning,” a term borrowed from the pragmatist philosopher Charles Sanders Peirce (Gillespie et al., 2024). Abduction is the logic of surprise. You encounter something unexpected, and you generate a new hypothesis to explain it. It is not deduction, which moves from general rules to specific predictions. It is not induction, which moves from specific observations to general patterns. Abduction is the creative leap that happens when the world refuses to fit your assumptions.

This has practical consequences. A pragmatist researcher designs studies that can accommodate surprise. They build in checkpoints where they can step back, reassess, and change course. They treat their initial question as a starting point, not a destination. They are willing to abandon a beautiful method if it stops serving the inquiry.

The Multi Resolution Solution

One of the most concrete contributions of Gillespie and colleagues is a new research design they call “multi resolution research” (Gillespie et al., 2024). The name is deliberately technical, but the idea is simple: you study the same phenomenon at different levels of detail, using different methods, and you let the results inform each other.

Imagine you want to understand why people join extremist groups. A single resolution study might survey 10,000 people about their political beliefs. That gives you breadth but no depth. Another single resolution study might interview five former extremists about their life stories. That gives you depth but no breadth. A multi resolution study does both, and more. It might start with a large scale survey to identify patterns. Then it might use those patterns to select a smaller group for in depth interviews. Then it might analyze social media data from the same individuals to see how their online behavior matches their self reports. Then it might return to the survey data with new questions generated by the interviews.

Each layer of analysis informs the others. The quantitative data provides the map; the qualitative data provides the terrain. The authors argue that this approach is particularly valuable for analyzing large sets of qualitative data, a challenge that has become urgent in the age of social media, online forums, and digital archives (Gillespie et al., 2024). Traditional qualitative methods were designed for small samples. Multi resolution research scales them up without sacrificing their interpretive power.

What Pragmatism Actually Demands of You

Gillespie and colleagues are not offering a free pass. Pragmatism is often misunderstood as a kind of methodological anything goes, as if the only criterion for a good method is that it works. But the authors are explicit that pragmatism imposes rigorous demands.

First, you must be clear about your goals. What are you trying to achieve? Is it prediction, explanation, description, intervention, or some combination? Different goals require different methods. A study designed to predict election outcomes looks very different from a study designed to understand why people vote the way they do. Pragmatism forces you to specify your purpose before you choose your tools.

Second, you must be accountable to consequences. The pragmatist test of a method is not whether it follows the rules but whether it produces useful results. This sounds permissive, but it is actually demanding. It means you cannot hide behind methodological orthodoxy. If your study fails to answer the question, you cannot defend it by saying “but I followed the protocol.” You have to go back and try something else.

Third, you must embrace uncertainty. Pragmatism does not promise certainty. It promises progress. The authors argue that research should be judged by whether it moves us forward, not by whether it settles the matter forever (Gillespie et al., 2024). This is a hard sell in a scientific culture that rewards definitive findings. But it is more honest. Most research questions are not settled by a single study. They are gradually clarified by a sequence of studies, each one refining the question and the method.

The Ethics of Pragmatism

The pragmatist approach has implications for research ethics that go beyond the standard checklist of informed consent and confidentiality. Gillespie and colleagues argue that ethics should be integrated into every stage of the research process, not just tacked on at the beginning (Gillespie et al., 2024).

This means asking ethical questions about your goals. Is this research worth doing? Who benefits from it? Who might be harmed? It means asking ethical questions about your methods. Are you treating participants as subjects or as collaborators? Are you extracting data from communities without giving back? It means asking ethical questions about your conclusions. How will your findings be used? Could they be weaponized?

The authors are particularly concerned about the ethics of big data. Large scale qualitative datasets, such as social media archives, raise questions about consent that traditional methods did not. When you analyze millions of public posts, are the people who wrote those posts participants or sources? Do they have a right to know how their words are being used? Pragmatism does not provide easy answers to these questions, but it insists that they be asked.

What This Approach Does Not Solve

Pragmatism is not a cure all. Gillespie and colleagues acknowledge that their approach has limitations.

One open question is how to handle conflicting results from different resolutions. If your survey says one thing and your interviews say another, which do you believe? The authors suggest that this tension is productive, that it reveals something about the phenomenon that neither method could capture alone. But they do not provide a simple rule for resolving the conflict.

Another question is about power. Pragmatism assumes that researchers can freely choose their methods based on their goals. But in practice, methods are embedded in institutional structures. Grant agencies favor certain methods. Journals publish certain methods. Tenure committees reward certain methods. A junior researcher who wants to use a pragmatist approach may find themselves swimming against a strong current.

A third question is about truth. Pragmatism defines truth as what works, but this can slide into relativism if you are not careful. The authors are aware of this danger and argue that pragmatism is compatible with a correspondence theory of truth, the idea that true statements match reality. But they do not fully resolve the tension. It remains an open debate within pragmatist philosophy.

These are not fatal objections. They are invitations for further inquiry. That is the point.

What This Actually Means

Here is the bottom line for anyone who does research, funds research, or consumes research.

• ▸Stop treating methods as tribes. You do not have to be a quantitative person or a qualitative person. You are a person trying to answer a question. Use whatever tools the question demands. If that means learning a new method, learn it. If that means combining methods, combine them. The only unforgivable sin is letting your methodological identity limit your inquiry.

• ▸Design for surprise. Build checkpoints into your research where you can step back and ask: Am I still asking the right question? Does my method still serve my goal? If the answer is no, change course. This is not failure. It is the whole point of research.

• ▸Think in layers. Do not settle for one type of data. Use surveys for breadth, interviews for depth, archives for history, experiments for causality. Let each layer inform the others. The most powerful insights come from the friction between different kinds of evidence.

• ▸Be honest about your goals. Before you start a study, write down exactly what you are trying to achieve. Is it description, explanation, prediction, or intervention? Different goals require different methods. If you are not clear about your goal, you cannot choose your method wisely.

• ▸Ask who benefits. Every research project is an intervention in the world. It takes time, money, and attention. It affects participants, communities, and institutions. Before you start, ask yourself: Is this worth doing? If you cannot answer yes, do not do it.

The pragmatist approach does not make research easier. It makes it harder. It demands more thinking, more flexibility, more ethical reflection. But it also makes research better. It produces work that is more useful, more honest, and more alive to the complexity of the world. That is a trade worth making.