Scientific Practices and Ethics
This information, and the package of general information is to be given to each new trainee entering the immunology graduate program. As of 2002 it is now provided on the Immunology website.

At the beginning of your scientific career, an explicit statement of appropriate "scientific ethics" is worth examining. Dr. Kent HayGlass prepared the information below and has given it out to new members of his lab since the late 1980's. Some of it is specific to his lab, so you should definitely consult with your supervisor to identify the lab practices (ie. concerning record keeping) that are used in your new lab.

Most practices are learned by example, and this is a topic that you will discuss with your supervisor, colleagues and others on multiple occasions. The following is prepared as a starting point from which you can consider issues likely to arise during your career – it is intended to act as a conversation starter than a manual.


Summaries of experimental design, raw and analyzed data can be kept in your binder in whatever fashion you find most efficient. A bound lab notebook book is not necessary as far as I am concerned. Your main record keeping should be on the computer, with a hard copy for your binder. Always maintain a backup copy of your computer files. (Different labs have different policies. Determine what method your advisor wishes).

Every experiment requires:

  1. A clearly stated purpose (Must be explicit and brief, not "we will look at...")

  2. A hypothesis that has been thought through. Most people find it useful to model the expected data while designing an experiment, and therefore see what additional controls need to be included for your interpretation to stand.

  3. Method - (which may just refer to previous experiments with the differences in this particular experiment highlighted. For the first time, all details should be identified very explicitly).

  4. Data - Microsoft Excel tables are the most organized and facilitate later statistical analysis. Figures should be created in Slidewrite or Prizm. Use common sense in analysing your data in addition to statistics. If your "identical" duplicates set up the same day are substantially different from one another (something you determine by chatting with your supervisor) then it doesn't really matter if your p value is <0.05 when comparing two groups. You have a problem.

  5. Specific conclusions you draw. Conclusions are usually followed by a statement about what needs to be done next. Each experiment must have a clearly, and explicitly, identified purpose, data and conclusions or you are wasting your time.

All raw data (tapes from counters, floppies with data from ELISAs, sheets used in injections, animal cards etc.) should be dated, marked with the number of the experiment involved and kept, though not usually in the binder. Your rough calculations and day-to-day notes should all be done on a steno pad. Never use paper towels, or scrap paper for ANY calculations because when the experiment fails - and some will - you won't be able to reconstruct why.

This data should be kept at least for a year or two after publication of your results in case they are challenged. Or, you may wish to re-examine them for something you might have missed when looking at them initially. If you wish, you can stack them in a drawer or box as, usually, you are unlikely to look at them again, so you do not necessarily need an extensive filing system.

You should label the raw data well enough that you can identify what was what, if necessary. (Which group is which, what was done etc - the information needed to reconstruct what you did) . Having the date, type of assay and Experiment number is usually sufficient. When you leave the lab, tell your supervisor where this information is kept. Different labs have different policies about who keeps the originals. It is most appropriate if everyone gets a copy.

Written by: Dr. Kent T. HayGlass
Last revision: October 14, 2002