In the world of paranormal investigation, one phrase is heard again and again and again — “we follow the Scientific Method.” It’s usually said with pride, sometimes as a boast and, for others, a badge of credibility and discipline.

However, there is a problem. Those who believe they operate using scientific methodology are, more likely without even realising it, claiming a process which simply does not fit the subjective and unpredictable nature of what we’re studying.

In all the years I’ve been involved with paranormal research, the more I think about it, the more convinced I am of this. The truth of the matter is simply you cannot reliably apply the current scientific method to the study of alleged paranormal phenomena.

This isn’t intended as a criticism of anyone’s work, it’s simply a reality check upon how we frame what we do.

So, what is the Scientific Method?
In its most fundamental form, the scientific method involves:

• Observation
• Hypothesis formation
• Controlled testing
• Analysis
• Replication by independent parties; and
• Peer review and communication.

At the heart of scientific methodology lies one crucial principle, replicability. If a phenomenon is both real and measurable, anyone using the same tools and methods should be able to observe it, repeatedly and, reliably.

And this, if you haven’t guessed it already, is where paranormal research runs into great difficulty.

Can you think of any alleged paranormal phenomena that would meet any of the above criteria, repeatedly, let alone all of them?

This is where other respected disciplines adapt.

In Archaeology, for example, absolute replication is impossible. You cannot “recreate” a dig site once it has been excavated. However, methods are modified to gather as much repeatable evidence as possible before variables change.

Psychology often faces unpredictable human behaviour, yet researchers use statistical tools and large sample sizes to account for this variability.

Anthropology also deals in unrepeatable, context-specific events, but relies on long-term observation and triangulation across multiple independent studies.

The difference is, in all these fields, the phenomena under study are still consistently accessible. In paranormal research, the very subject of study can vanish without warning.

Why paranormal phenomena don’t fit the model
To demonstrate this, let’s use EVP (Electronic Voice Phenomena) as an example.

You can record in the same location, using the same equipment and ask the same questions, yet get wildly different results.

Even if an anomalous voice is captured on a recording device, there is absolutely no guarantee that anyone else (or even yourself, during the same session if you’re using more than one recording device) will be able to replicate it under exactly the same conditions.

This lack of repeatability breaks a fundamental rule of the scientific method. And, sadly, it doesn’t apply to just EVP.

Apparitions, the movement of objects, cold spots, responses through devices are just a few examples I could use. All of these are inconsistent, spontaneous and yet appear contextual and intelligent in ways that cannot be reproduced at will.

So, this leaves us with the question, what do we do with a phenomenon that may be influenced by emotions, intent, timing, environmental factors or even consciousness itself?

We cannot isolate and control these variables. Indeed, it’s true to say that sometimes we cannot even identify them.

Some may argue that controlled laboratory experiments, for example, setting up environmental chambers to simulate haunted conditions, could solve this.

While such setups may reduce environmental noise, they have yet to consistently produce phenomena under observation. In fact, there’s a danger that moving the study into a laboratory strips away the very factors, such as historic and human emotional context and, the environmental interplay that may be central to the occurrence in the first place.