Apply Six Scientific Principles to UX Research

There are 6 scientific principles that are foundational in scientific thinking. Understanding what they are can help perform more focused research.

Note: While I apply these 6 principles to UX research, I believe they are applicable to many industries.

The 6 principles:

  1. Rule out rival hypotheses
  2. Correlation isn’t causation
  3. Falsifiablity
  4. Replicability
  5. Extraordinary claims need extraordinary evidence
  6. Occam’s Razor

One: Ruling Out Rival Hypotheses

Before reaching a conclusion — before or after research — make sure that all other explanations have been considered & vetted.

In-practice: Reach out to other teams that interact with your users & research topic and understand their data and approach. Chances are there is some similarity that help focus your efforts.

Two: Correlation Isn’t Causation

While heard by many, this principle forces researchers to understand the relationship of their user data: No single cause can account for all changes between variables.

In-practice: When engaging with stakeholders, avoid stating direct causes within user data — explain how human behavior is influenced by many factors, including the tested product.

Three: Falsifiability

If a hypothesis cannot be proven false, it is not science. All research hypotheses must be able to be incorrect.

Example: Apples are friends with bananas. [This is a poor research hypothesis because this statement cannot be proven to be incorrect.]
Example: Bananas have more potassium than apples. [This research hypothesis is valid — because it can be proven wrong — and offers a more concrete research direction.]

In-practice: Try your best to engage in impactful & specific user research, instead of research questions based on wild assumptions, or gut feelings.

Four: Replicability

Can the results be found in another duplicate study? If not, then the research method is potentially faulty.

In-practice: Test with numerous representative users! Test with concepts, test with early iterations, test post-release. Note the results and how the successes and failures change over time.

In an ideal world, replicability is best done with many researchers testing the same research question with representative users.

Five: Extraordinary Claims need Extraordinary Evidence

Is the evidence as strong as what is being claimed? The more outrageous a claim, the more outrageous and concrete the data must be.

In-practice: Any statement in a research document or presentation must be backed by strong evidence. Use results boldly — but back them up with numerous instances of data.

Six: Occam’s Razor

Favor simple explanations over complex ones. More often than not, a basic understanding arises from data, rather than hidden relationships.

In-practice: Tell the story of the data & and your users; don’t extrapolate deeper meaning. If you believe there is a more intricate link between your users and the product, look to study it.

UX research is science. Applying scientific principles to user research can only increase its strength and its effectiveness.

Check out to learn more about me (desktop-only).