Planned vs. Discovered: What’s the Difference?
In the world of research, there are two main types of discovery: confirmatory and exploratory. Think of it like a road trip. A confirmatory study is like planning a precise route from Mumbai to Delhi. You have a map, a specific destination (your hypothesis),
and you are testing whether you can get there as planned. The goal is to confirm or reject a pre-existing idea. An exploratory study is like getting in the car with no set destination, just a desire to drive and see what interesting landmarks you might find along the way. You might discover a fascinating village you never knew existed. Both types of journeys are valuable. The planned trip gives you a reliable answer to a specific question, while the exploratory trip generates new, exciting ideas for future trips. The problem arises when the exploratory discovery is presented as if it were the planned destination all along.
The Danger of Painting the Target Afterward
Presenting an unexpected finding as if it were predicted from the start is a questionable research practice known as HARKing, which stands for “Hypothesizing After the Results are Known.” Imagine a sharpshooter firing at a barn door and then drawing targets around the bullet holes. That’s HARKing. It misrepresents the research process, making a surprise discovery seem like a confirmed prediction. This often goes hand-in-hand with “p-hacking,” where researchers might analyze their data in various ways until they find a statistically significant result, even if it's just due to chance. These practices increase the risk of false positives entering the public conversation. An exploratory finding is a clue, not a conclusion. It needs to be tested with a new, planned (confirmatory) study before we can be confident in it. When that distinction is blurred, we end up with a literature full of seemingly positive results that are difficult to replicate.
The 'File Drawer' Problem
Journals and media outlets are often more excited to publish splashy, positive results than studies that find nothing. A study that sets out to prove a hypothesis but fails is called a null result. Too often, these perfectly good studies end up unpublished, tucked away in a researcher’s “file drawer.” This creates publication bias. If 20 studies are conducted on a topic, but 19 find no effect and only the one that found a surprising link gets published, our view of the evidence becomes heavily skewed. This makes it seem like a single, possibly random, finding is the whole story. It exaggerates the number of positive results in circulation and makes it harder for the scientific process to correct itself.
A Push for Transparency: Pre-registration
So how do we fix this? The strongest solution being adopted in the scientific community is pre-registration. Before collecting any data, researchers write down their plan—their hypothesis, methods, and analysis strategy—and post it to a public registry. This time-stamped document acts as a safeguard. It doesn’t stop researchers from making unexpected discoveries during their work, but it ensures a clear line is drawn between what was planned and what was discovered along the way. This simple act of transparency prevents HARKing and p-hacking, as anyone can compare the final published paper to the original plan. It also encourages the publication of negative results, helping to solve the file-drawer problem and giving everyone a more complete picture of the evidence.
