The paranormal continues to baffle after well over a century of serious investigation. One of the reasons is that the mass of data collected simply doesn't make much sense when you try to analyse it. This either means the paranormal is unexplainable, as some maintain, or that the data contains too much spurious and misreported material. It soon becomes obvious to anyone studying the data that the latter is likely to be correct. The standard of information ranges from papers in scientific journals to 'friend of a friend' anecdotes. Worse, in the absence of investigation standards, it is impossible to tell which reports are reliable and which spurious.

This whole problem could be summarised as 'misreporting'. It is endemic to the field for reasons explained below. Unless investigators recognise the problem they will continue to rely on very mixed data. For instance, contrast the picture of ghosts drawn from careful investigations with the popular view, demonstrated in TV ghost hunting shows. This article examines in detail how misreporting happens in paranormal research and how it can be avoided.

Witnesses

When someone comes across an apparently paranormal event unexpectedly, their evidence as a witness is valuable because they are not subject to psychological suggestion. Unfortunately, it is also problematic. Witnesses to unexpected events routinely misreport what they've seen. Former ASSAP Chairman Phil Walton did extensive experiments on witness testimony over many years (see here). He discovered that when a group of witnesses saw a staged event, in plain view in good lighting conditions, they gave a surprisingly wide range of answers when asked to describe what happened and who took part. With single witnesses (common in paranormal cases) the problem is even worse because you can't take an average of the reports.

Viewing conditions for people seeing ghosts are often far from ideal. Also, witnesses may find the sight of a human figure vanishing shocking, further compromising their ability to accurately recount events. Such misperception is a serious problem with paranormal reports.

Human memories are not formed like recording data being stored on a computer disc. For a start, most people cannot recall an event with photographic clarity (eidetic memory). Some bits, that are not properly stored, may be 'filled in' with details from imagination when recalled later. Once such false details are 'recalled', they become part of the long term memory of the event, feeling as real as the rest. This can happen, for instance, if you ask a witness to recall the colour of an object when they don't really remember it. Having unconsciously decided an object was red that will remain part of their memory of the incident, even though the object was really green. Sometimes major components of the memory of an event can be replaced by ' false memories' in this way (confabulation).

Memory can also be affected by prior belief (suggestion). Imagine the situation where you see a fleeting shadow in a dark corridor. It might be caused by a curtain fluttering in the wind but, if the place is known to be haunted, your mind might 'decide' it is really a ghost. The 'decision' (essentiality about what is reality) is taken by your brain before it even reaches your conscious mind (which is why it is so convincing). So, when later questioned, instead of reporting a fluttering curtain, you might describe a human figure with convincing details such as clothing colours, hair length and approximate age! It is important to find out witnesses' attitudes to the phenomena they have witnessed, in case it introduces bias.

Recording events

Witness testimony is, as we've seen, not hugely reliable. However, it is frequently our only access to paranormal events so we must work with it, while acknowledging its limitations. This vital information, whether accurate or not, is now in the witnesses head, so they need to be interviewed to extract it.

Interviewing is a skilled process. It is easy to adversely influence the outcome of an interview by using such inappropriate techniques as leading or closed questions. Leading questions suggest the answer that is desired while closed questions offer very few alternative answers, all of which may sometimes be wrong! A common mistake of some interviewers in paranormal research is to bias their questions according to their own prior beliefs. They might emphasize questions that support their beliefs while omitting others that might suggest alternative explanations. As discussed above, such belief orientated questions may actually affect the permanent memory of the witness. If the interviewer starts talking about 'spirits', it is unsurprising that the witness will start to remember their 'ghostly encounter' as one with a 'spirit', rather than simply seeing a vague figure in a hallway (which was actually a curtain!).

Finally, once an interview is recorded in written form, it will inevitably be incomplete and sometimes inaccurate. There is little that can be done about this but it must be taken into account nevertheless. A detail that might have proved that the sighting was xenonormal may be omitted because no one thought it relevant at the time of the interview.

As well as interviewing witnesses, investigators would normally examine the site of the incident. This might reveal, for instance, that things could not have happened precisely as the witness said because of what can, and cannot, be seen from where they said they stood. Site details also make it much easier for anyone later examining the report to appreciate what the incident really looked like. We all have a tendency, when told of an incident, to form an involuntary imaginary picture of it in our heads. These imaginary pictures are almost invariably wrong but, nevertheless, can affect how we interpret the report. We might, for instance, be impressed with a ghost sighting until we realise that there is a fluttering curtain, that the witness failed to mention, in the place where the ghost was supposedly seen.

Some paranormal researchers give very little attention to interviewing witnesses, being keener to get on with a vigil. This is despite the fact that the witness may well be the only person in the entire investigation who actually sees a ghost!

Second-hand information

Some reports of paranormal phenomena arrive with researchers second hand. For instance, the original witness may not available to be interviewed for some reason. Or investigators may rely solely on what previous investigators have said. Other sources of such 'second-hand' information include newspaper reports, popular books, 'friend of a friend' stories and even TV programmes. Such reports are effectively anecdotal, as you may have no idea how they were collected, and should be treated with extreme caution.

You might think that if you take a lot such anecdotal information together, the differences might even themselves out, so still allowing us to produce an accurate overall picture of ghosts. This is true for certain sorts of data but not paranormal phenomena reports. There are common beliefs, like the one that ghosts are spirits or ghosts are recordings which will massively bias the data (see above) in an unpredictable way. For instance, investigators who are convinced that ghosts are spirits might miss obvious clues that apparitions were caused by sleep paralysis or hypnagogia. Given the the highly variable standard of second-hand data collection, such unpredictable biases are too serious a problem to make such reports useful. It would be dangerous to base future investigations on such data, or to use it to come to conclusions about paranormal phenomena.

Technical misunderstandings

The paranormal is defined as 'beyond the normal'. Therefore, to demonstrate that something is paranormal, you need to eliminate all normal explanations. This relies on the investigator understanding the technical aspects of all possible natural explanations. Unfortunately, we cannot all be experts in everything so, inevitably, some phenomena are labeled paranormal when plainly they are not. A good example is orbs. All sorts of objections have been raised to the idea of orbs as photographic artifacts in an attempt to show they are paranormal (see a list here).

It is such technical misunderstandings that lead to many reports of paranormal phenomena, particularly on vigils, which are really xenonormal. Just as interviewing is a skill, so is understanding technical data.

Assumption-led investigations

In recent years, many people or groups have started to use assumption-led investigation methods. Such methods are inevitably biased and tend only to reinforce their own assumptions. For instance, many such investigations involve the use of mediums. In contrast to instruments, whose function, accuracy and limitations are well understood, we know none of these things with mediums. Mediumship is a worthwhile area of study on its own but bringing mediums into investigations increases the unknowns in the situation when the object should be to reduce them. Given the bias and unknowns involved, any results from assumption-led investigations should be treated with extreme caution.

Before the rise of the TV ghost hunting shows there were few people researching ghosts. Interestingly, they turned up little, if any, evidence to support the idea that ghosts were spirits. Instead, many favoured the stone tape theory. Since the TV shows, interest in ghost research has expanded hugely and the idea of 'ghosts as spirits' is practically a given among the newcomers. Has the evidence from cases changed with the arrival of new investigators? That hardly seems likely. What HAS changed is the way many cases are investigated. They no longer start without assumptions about the nature of ghosts.

Birth of a legend

We are all familiar with the idea of Chinese whispers, where a story is distorted by successive retellings. This happens as much in paranormal research as anywhere else. However, it causes a bigger problem in the paranormal because of the in-built biases common in those doing the retelling.

Newspapers, for instance, like to have an 'angle' on a story. So, if someone reports seeing a 'figure' near an old building, a reporter may dig around in the archives to produce a 'suspect' (someone who once inhabited the building) that might 'account' for the ghost. This can happen even when the figure doesn't resemble the suspect, or indeed anyone at all. Soon the ghost becomes associated with the suspect in popular imagination and subsequent news stories and books (see Blue Bell Hill ghost as an example).

Popular regional collections of true ghost stories are frequently published in book form. Many of the stories featured are so old that they may be little more than legend. Whatever original incidents inspired the stories have been hugely exaggerated by retelling. Some of the people and incidents in such stories may not even be traceable in historical records. For an example of this process see here. Unfortunately, sometimes these ancient stories, though clearly exaggerated, are taken seriously by modern researchers.

Fixing the problem

Given the major problems that paranormal research has with misreporting, are there any remedies? Luckily, there are!

Witnesses: With witnesses the remedies are: (a) get multiple, independent witnesses and (b) ensure they are interviewed carefully. Independent witnesses are ones who know nothing about each other, or the events they see, beforehand. Also, they should not be deliberately seeking paranormal phenomena because that introduces the possibility of psychological suggestion. If such independent witnesses produce similar accounts of events (at the same place but at different times in the case of ghosts) then there is probably something worth investigating.

Second-hand information, assumption-led investigations and legends: Ideally, this data should be rejected. If it IS used, a health warning should be given that it is indicative only.

Technical misunderstandings: Simply ensure that people with relevant technical skills are available for investigations.

Conclusions

Misreporting is a serious problem in paranormal research that is rarely acknowledged. Too many people draw conclusions from surveys of data which is little better than anecdotal. It is, therefore, no surprise that many myths continue to propagate around the field. The persistent ideas that 'ghosts are spirits' and 'orbs are paranormal' are, despite popular belief, not supported by the best data. And yet, it is just such assumptions that determine the course of many investigations. Thus, sadly, much useful data that might have otherwise been gathered is forever lost.

© Maurice Townsend 2008