Exploratory vs Confirmatory Research

Are you familiar with "exploratory research"? Did you know it's one of the most effective tools for enhancing efficiency in scientific endeavors? And are you aware that the tools it requires are fundamentally different from those utilized in confirmatory research?

Today, I aim to delve into these topics.

Two Complementary Approaches

Let's start by defining these concepts succinctly:

• Exploratory research is a phase in the research process designed to link ideas and uncover the "whys" behind potential cause-and-effect relationships. This phase kicks off when researchers begin to decode what they are observing while developing cause-and-effect models.
• Confirmatory research, also known as hypothesis testing, is conducted when researchers have a solid hypothesis about what is transpiring. Here, the researcher possesses a theory (or multiple theories), with the goal being to determine if empirical evidence supports this theory.

How Do They Complement Each Other?

The crux of the matter is that exploratory and confirmatory research serve as two integral parts of a singular objective: to unearth significant findings in the most efficient, reliable, reproducible, and applicable way possible. 

These approaches are complementary because research always initiates with a complex, unknown issue - an issue that needs observation even before determining which hypotheses require further validation.

As highlighted by J.W. Tukey in his influential work, we need both exploratory and confirmatory methods because:

“Finding the question is often more important than finding the answer”.

A pertinent question now arises: should exploratory analysis be concerned about false positives? To shed light on this, consider the perspective of Andrew Gelman, a political scientist and statistician at Columbia University in New York and author of one of the most esteemed blogs on this subject. In a recent article for Nature magazine, he states:

"… In this approach, exploratory and confirmatory analyses are approached differently and clearly labelled. … Researchers would first do two small exploratory studies and gather potentially interesting findings without worrying too much about false alarms. Then, on the basis of these results, the authors would decide exactly how they planned to confirm the findings.”

Therefore, the task of exploratory analysis is to unearth potential relationships and generate new, pertinent questions that enrich the traditional confirmatory process aimed at minimizing type II errors - that is, the oversight of existing phenomena.

Want to Read More?

• Fruitful Collaborations: real success cases at applying automated EDA.  
• How To Design an Efficient EDA Study : the 3 key ingredients of the EDA recipe.  
• 5 Kinds of Exploratory Questions You May Be Asking Yourself : from our experience ...  
• Biomedicine and Complex Data : the concept of "complex data" and its implications in the biomedicine research field.