How to Write Up a Night
Everything you did that night stops existing about a week later unless somebody wrote it down in a shape that can be read cold by a stranger — which within a year includes you.
The Survey's report format has not changed materially since 1974, and the reason is not tradition. Every section in it exists because a file was once useless without it.
The format
Site. Precise enough to return to and no more precise than the occupier permits. Where somebody has asked for the address to be withheld, that is recorded on the file as a decision, not as an omission.
Date and personnel. Who was there, and their designation. A file with an anonymous observer in it cannot be followed up.
Conditions. Weather, outside temperature, occupancy, heating, what was running, what was switched off and by whom. Half of the ordinary explanations available to you a year later are in this section, and none of them are recoverable if it is missing.
Instruments carried. Every item, with its revision or serial. If a figure is ever queried, the first question is what produced it.
Baseline. Positions, values and ranges. See the note on baselines; this is the section that makes every other number mean something.
The log. Timestamped, written at the time, in the order things happened.
Analysis. What was reviewed afterwards, what was attributed to what, and what was not.
Determination. One of two sentences: cause established, followed by the cause; or cause not established. There is no third option and there is no space for a hedge.
The log is the difficult part
A log entry is a time, a position, an instrument, a figure and a flat sentence. It contains no adjectives and no conclusions.
22:41 — R3, position 2, Instrument 01, 4.8° (floor 1.2°). Held nine seconds, fell over about two. Operator stationary. Nobody else on this floor.
Not: the meter suddenly spiked and it felt like something walked past. The first version can be checked, plotted, compared with a recording and argued with. The second cannot be anything except believed or disbelieved.
Three habits make the difference:
- Write it at the time. A log reconstructed at 4am is a memory of a night, and memory reorders events by how significant they turned out to be. That is exactly the transformation you are trying to avoid.
- Log the boring things too. The car door, the boiler firing, somebody going to the toilet, the moment you moved the tripod. These are what let you rule things out later, and they are invisible in a log that only records the interesting parts.
- Log what you did, not only what happened. Half the events in an amateur log are the logger, and there is no way to know that unless the logger wrote down where they were.
Where the feelings go
They go in the report. They do not go in the log.
The Survey's operators have been told for fifty-two years to keep subjective impressions out of the log, and for fifty-two years they have put them in the margins instead, and the archive is transcribed as written, so the margins are in the files. Read a few of the case files and you will find them, and you will find that they are frequently the most memorable thing on the page.
That is not an argument for mixing them in. It is an argument for the format. The reason a pencilled line in a margin lands as hard as it does is that everything around it is a number.
Writing the determination
This is the section people avoid, and the avoidance is always the same shape: a paragraph that describes what was found, gestures at what it might mean, and never commits.
Write the sentence. If the cause was a fridge, the determination is cause established: the compressor of the kitchen refrigerator, duty cycle approximately four minutes, and the file is closed and is worth exactly as much as any other file. If you did not work it out, the determination is cause not established, and that is a real finding, and it is only worth anything at all in an archive where the first sentence is common.
A group whose files all say not established has not established anything, including that.
Keeping it
Three practical things that cost nothing and save the whole record:
- One file per site, not per night. The second visit to a building is where the value is, and it is only accessible if the first one is filed where you will find it.
- Keep the raw material. Original recordings, unprocessed. Not just the clips. A processed clip cannot be re-reviewed and cannot be handed to somebody else blind.
- Write the site's ordinary facts down once. The plan, the position of the consumer unit, the fridge cycle, where the draughts are. In two years that page will be worth more than everything else in the folder.
What this station keeps for you
The instruments here file into a service record held in your own browser and nowhere else. It stores measurements rather than sentences — the value, the floor, the sigma, the bearing, the duration — because the wording is a rendering decision and the numbers are the part somebody wants in six months.
You can name the room you are working in, and the label stays on your device. That is the feature worth using, and its point is the second visit: a room whose floor was 1.83° in March and 1.83° in November is a room you now know something about. It is the only honest version of is this place haunted that a website can offer, and it takes two visits and a name to get it.
Questions
- What should a paranormal investigation report include?
- Site, date, personnel, conditions, equipment carried, the baseline, a timestamped log, the analysis, and a closing determination that states plainly whether a cause was established. Nothing about what anybody felt belongs in the log; it belongs in a separate, clearly marked section.
- How soon should you write up an investigation?
- The log is written during the night, at the time, not afterwards. The analysis is written within a week, while the recordings are still fresh enough to be re-checked and the people involved can still be asked what they meant.