As I examined some of the data, I noticed that there were several
plots that contained one or two isolated points at a great distance
away from the mean. This occurred when I got data with a high number
of photon counts (i.e. low rates). Most of these points were much lower than the
mean value. I turned to statistics to analyze whether these counts
were feasible or not. Statistics tells us that for a normalized
distribution, 99.994% of the data lies within 4 standard deviations
() from the mean. For a non-normalized distribution, the same
percentage lies within about 7 standard deviations from the mean.
I ended up choosing 7 standard deviations to be the limit for whether
I should include the counts or not in my data. Therefore, anything
that fell outside of this range I interpreted as dropped counts due to
experimental error or error from the equipment.
Here is an example of one of the plots where I saw a dropped
count. You can see one data point at about 2.5 seconds after the
start of the experiment that is isolated
from the rest of the counts.