Keeping a lab notebook
In all experimental physics it is essential to keep adequate notes. In research scientists always use bound notebooks, not loose-leaf ones, for recording details and experimental results. The school policy is that you must buy at least three A4 bound notebooks, two for ordinary experiments and one for projects. The type with millimetre graph paper on the left hand page (e.g. Chartwell Laboratory Book type A4 641-C, or A4 642-K, or Greenleaves SR8A4) is obtainable from Stanford and Mann or the Guild shop. A suitable notebook may be purchased from the laboratory at a cost of £1. Linear graph paper is freely available during laboratory sessions.
These notebooks are to be your diary of everything that you have actually done in your experiments. It is not intended for "writing up" experiments.
A good account of laboratory techniques and how to keep a laboratory notebook is given in chapters 10 - 12 of Squires (see other resources below).
The crucial test of a good notebook is that it provides you (not the demonstrator) with a complete record of everything that you have done so that all relevant facts can be recalled, perhaps even after a substantial amount of time has passed.
There have been a series of law cases about who should have the income from a patent on all the HAS lasers being made. Many companies have been sued and some have gone out of business. The crucial evidence was one man's record of three days work as recorded in his laboratory notebook in 1961. Imagine losing millions of dollars of income because you never got into the habit of keeping good records! Incidentally, under US law a laboratory notebook, properly dated and signed by witnesses, is accepted as establishing priority for patent applications. At the time of writing, that is not the case in the UK.
Starting a new experiment
- Read the details on the relevant experiment before starting work, preferably the day before your laboratory class.
- Write down the date, experiment name and bench number. You should also write down the date at the start of each day's work.
- Plan the experiment. Become familiar with the apparatus to see how it works. Record preliminary results to give you an idea of the range of readings to be expected. This will also help you to have a sensible layout on the bench so that equipment that needs to be adjusted carefully can be close to your best hand.
- Think about likely sources of error. Identify the main ones and decide how you are going to deal with them.
- Plan how you are going to record data. For example, if you were looking at the kinetic energy of a fluid you would need to record velocity and then calculate (velocity)3, so you would need to leave a column in your notebook for the calculation. Don't accumulate large amounts of unprocessed data as this will only lead to disaster.
Doing an experiment
- Write observations down in your book while you take them. Don't use bits of paper because these can be lost or taken away by a demonstrator.
- Record actual data not derived data. So for the velocity example in point four above, you would record the velocity and then calculate the cube, not calculate the cube and only record that. This can be useful in tracing computing or calculator errors.
- Draw graphs as you take the data. This lets you see how the experiment is progressing and spot any 'wrong' points due to misreading the equipment. It doesn't matter if the scales of the graph are wrong, because a better graph can be drawn later - what we want here is to be alerted to mistakes in procedure while the equipment is still set up.
- Record everything important, e.g. circuit diagram, even if it is in the manual. Record the voltages measured. Record the zero error of equipment (even if it was zero). Also it's a good idea to record apparatus numbers - this way you can be sure the inductor you are using this week is the same one you carefully calibrated the previous week!
- Be neat so that you can find your way through afterwards - neatness helps to avoid errors.
- Spread out your notes. Leave spaces so comments, calculations or corrections can be added later if needed.
- Be as brief as possible, but don't omit units on tables of results or graphs. There's no need to write sentences or perfect prose.
- Don't reproduce the manual, unless, for example, you find it useful in understanding the theory, or if a diagram in the manual helps to state what was measured.
- Don't remove results that later prove to be wrong: put a single line through them and add a note stating why they were wrong. Being able to notice, and correct, one's mistakes is the sign of a reliable experimenter! Don't use correcting fluid.
- Always describe what you saw, as well as recording the numbers. Qualitative observations are just as important as quantitative.
At the end of an experiment
- Your notebook will also contain the analysis of results and any error calculations. Compare your result with accepted values from data books, if possible, and attempt to explain any discrepancies.
- A short verbal conclusion should be added, with comments on how the experiment went, main difficulties, major sources of error, and possible improvements etc.
A book which gives a good introduction to physics laboratory practice (but not related to the experiments you will do) is:
- Squires G.L., Practical Physics, 3rd edition (Cambridge University Press).
In addition, you should use a good book on error analysis, e.g. one of the following;
- Barlow, R.J., Statistics (Addison Wesley).
- Lyons L., Data Analysis for Physical Science Students (C.U.P).
- Taylor, J, R., An Introduction to Error Analysis (University Science Books).