Learn · Qualitative methods
Deductive, inductive, and abductive reasoning
If these three words never felt intuitive, you are not alone. The concepts are simple. The labels are not. Here is each one in plain terms, with an example, and why mixing them up weakens your argument.
After this guide you will be able to:
- Name the reasoning you are actually using.
- Match your sampling and coding to that reasoning.
- Spot when a paper claims one and does another.
The three, side by side
| Reasoning | In plain terms | Example | Use it to |
|---|---|---|---|
| Deductive | Top-down. Start with a theory, then test it. If the premises hold, the conclusion follows. | Theory: social support lowers stress. Study: do people with high support score lower on stress? | test or confirm an existing theory |
| Inductive | Bottom-up. Gather observations, find patterns, build a theory. The conclusion is probable, not certain. | Interviews produce categories, then themes, then a theory of the phenomenon. | build theory when little is known |
| Abductive | Best explanation. Move between theory and data, adjusting as you go. Plausible, not watertight. | A doctor reads the symptoms, infers the most likely cause, then tests it. | explain a surprising pattern |
Abduction sits between the other two. You may start from established concepts, but you have to be willing to adjust them, or invent new ones, when the data push back. That back-and-forth is what bridges theory and emergent insight, and it is also why abductive conclusions stay tentative until other evidence backs them up. (You sometimes see it miscalled "adductive".)
Why it matters
Deduction tests, induction builds, abduction explains. Use one word while you are really doing another, and the logic of your argument blurs. That is how a paper ends up making claims its design cannot support. For example, do not claim "code saturation" in a reflexive thematic analysis — saturation assumes codes exist to be exhausted, and that method rejects the assumption. The word has to belong to the method.
If the words were the hard part, that is normal. The reasoning here is something you already do. A detective, a doctor, a mechanic all move between theory and evidence every day. The difficulty is almost never the idea. It is that the labels are unintuitive and rarely explained. Once the word is yours, the logic snaps into place.
Keep the reasoning and the words aligned.
You called your study "inductive", and a reviewer pointed out that you started from an existing framework. These terms overlap in practice and travel loosely between traditions, so the mismatch is common. Name the reasoning you used and keep the methods section consistent from protocol to write-up. MethodVahti flags places where the stated method, terminology, and claims diverge. You still decide which reasoning fits the question.
See MethodVahtiRelated guides
See also code vs meaning saturation and the three questions that make qualitative research rigorous. More guides are on the Learn page.