The importance of keeping a good experimental logbook.

navigation page

Why is the experimental logbook important?

• It enables a complete reconstruction of the experiment or measurement at a later date.
• The "later date" can be many years; even after the death of the experimentalist.
• It enables the work to be repeated for re-evaluation of the reported results.
• The steps that led to the success or failure of a large project can be extracted.
• Patent lawyers need properly documented evidence of inventions.

What kind of book for a logbook?

• A bound notebook with numbered pages is essential so that alterations, removals, and additions, are difficult to make at a later date.
• It should have high grade archive quality paper.
• The notes should be legible, in non-fading ink, and written as the measurement proceeds, not afterwards from memory of what happened.

Truth in reporting observations and experiments

Results should never be massaged to improve their fit with preconceptions. Negative results are as important as positive results. Facts should be recorded without comment, opinion should not feature at all. Uncertainties should be noted, unrepeatability recorded.

What should be recorded in the logbook?

In general, anything which allows someone else to repeat what YOU did in the experiment.

• Steps and procedures which are not obvious
• Precautions taken to minimise error
• References to other people's work, ideas, hints, and inputs.
• Parameters which might affect the outcome of the experiment.
• Equipment used, type numbers, serial numbers, any calibration steps taken.
• Sketches of experimental layout and traces on recorders, oscilloscopes, spectrum analysers etc.
• The date and time
• The names of other people observing.
• Rough error analyses taken during the experiment, repeat observations of doubtful readings, calibration errors allowed for.

Data recordings

Raw data should be recorded as well as processed data, so that algorithmic error can be spotted at a later date.

You should aim to process and plot the data point by point AS YOU TAKE AND RECORD IT. This is a very important habit to acquire as it is often the quickest way to spot systematic error. You will find that the time taken to do this is repaid in time saved by not taking great strings of erroneous data.

Charts, photographs, and printer output from the experimental apparatus should be cut and glued in the logbook wherever possible. Labelling of data plots is very important. The originals should be used wherever possible. The axes should be labelled and each item should have a descriptive caption. Date and time should be recorded.

Validation of the logbook.

The logbook should be shown to someone else on the same day that it is written. The experimentalist and the observer should both sign and date the logbook. The observer should add the words "read and understood". This is important for patent law purposes. See BTG's pages and links therefrom for the patent law requirements for log books.