The Paranormal Database

[vbloke]

http://www.paranormaldatabase.com

could be a useful resource...

[Dr B]

all the main info on cumbrian ghosts is inaccurate.... >:(

[tkingdoll]

Is there a paranormal wiki, I wonder?

[Admin]

Is there a paranormal wiki, I wonder?

I hope not!

I have a feeling it would be an addictive read.

[Jocky]

As somebody who works professionally with proper databases, this kind of woolly pseudo-data makes me cross. >:(

There are a lot of people who are so blinded by the word "database" that they think if it's in one then it must be true. I am deeply skeptical of all such claims

[Admin]

That's a good point Jocky, people do seem to place weight on anything that's held in a database.

Of course data is of little use until it's processed into information and even then can be wrong.

I love gaining these little insights into how we can take things for granted when shouldn't.

[Jocky]

An example of the low value of this "data":

Cavalier
Location: Banbury - Reindeer Inn
Type: Haunting Manifestation
Date / Time: Unknown
Further Comments: Having vanished for many years now, it is reasonably certain this once gruesome entity has now gone forever.

I happen to know the Reindeer - an attractive coaching inn of 17th century date located in central Banbury, which does good real ale. However, if I hadn't darkened its doors I would have no real idea from this entry of where it was, as the "data" doesn't locate it accurately - no OS grid refs or postcode, which are the usual 'hard data' ways of identifying locations. Therefore the data can't be used to study distribution, beyond county boundaries - and even that seems to rely on the accuracy of text data entry, as the form doesn't have a county list to choose from.

Date unknown. Very helpful, that. No chance of chronological queries either. Entering the date seems to be optional on the form.

"Type: Haunting Manifestation" looks promisingly like a sound typological field which one could at least use to classify by type. But no, the entries appearing in the data don't seem to match the types available on the data entry form :( >:(

The "further comments" text field is likely to be worthless for proper querying, and in this case it even tells us that the alleged phenomenon no longer exists. Can we filter past phenomena from present?

Another thing: where is the field that tells us where the data came from?? No source - no confirmation.

Useless, as far as I can see. But credulous folk might point to this and say "Look, there are [insert spurious number here] of sightings properly recorded in a database - so it must be true". This is of course fallacious - Argumentum ad numerum and Appeal to Authority.