To prove that a phenomenom is real, usually three individual experiments are done, and each experiment is done in triplicate (three repeats), for a total of nine repeats.
Three experiments ensures there are no variations *between* experiments (such as the temperature/humidity on a particular day, contamination of your sample...). Therefore , these three experiments cannot be done in one setting; you'd have to do it on consecutive days. Idealy, the samples for these three experiments should come from different sources (tissues have to come from individual rats, cells have to be isolated from different patients....)
The repeats are done to insure there is not variation *within* each experiment (for example, results of pipetting error). If one of the repeats is markedly different from the other two, you'd have to do the experiment all over again. Some people do quadruplets (four repeats) so they can "kick out" the offending one.
Edit: btw, the standard deviations you see on graphs and tables are the std of the three experiments; the repeats are simply averaged.
Comments
To prove that a phenomenom is real, usually three individual experiments are done, and each experiment is done in triplicate (three repeats), for a total of nine repeats.
Three experiments ensures there are no variations *between* experiments (such as the temperature/humidity on a particular day, contamination of your sample...). Therefore , these three experiments cannot be done in one setting; you'd have to do it on consecutive days. Idealy, the samples for these three experiments should come from different sources (tissues have to come from individual rats, cells have to be isolated from different patients....)
The repeats are done to insure there is not variation *within* each experiment (for example, results of pipetting error). If one of the repeats is markedly different from the other two, you'd have to do the experiment all over again. Some people do quadruplets (four repeats) so they can "kick out" the offending one.
Edit: btw, the standard deviations you see on graphs and tables are the std of the three experiments; the repeats are simply averaged.
To get insight. It may reveal a trend. If a trend is revealed, then the experiment(trial) is repeated many times to conclude the results.
Prior to analyzing, what a scientist thinks is a mere hypothesis. After repeated experiements, it is a conclusive result.
because 20,000 rows of numbers isn't very informative
for analysis