I've been trying to determine the Boltzmann constant by observation of Brownian motion. I undertook four experiments and hence got four different estimated values for kB. To analyse the data, I estimated the mean value and the standard deviation. This is maybe not the best way to analyze the data since there are so few data points. I've heard of a method called "fractional uncertainties", but I was unable to find anything relevant by googling. What does the method consist of, and should I use it?