Physics307L F09:People/Joseph/Lab3

From OpenWetWare
Jump to: navigation, search

Summary for lab on the Poisson distribution

The point of this experiment was to show that under certain circumstances the probability distribution can take on a special shape, known as a Poisson distribution. A Poisson curve best fits data when there is a low probability of occurrence. We collected data by setting up a detector and recording incidences of cosmic radiation on the computer. Our collected data (of cosmic origin) didn't really fit the Poisson curve as well as we expected. There was a substantial amount of error, both random and systematic. We believed that some error is due to the fact that the experiment measures data of an uncontrollable origin, therefore we have no way of knowing when massive fluctuations may occur. We weren't terribly happy with our results because our data (found [here]) did not fit terribly close with the Poisson distribution. Our data wasn't perfect, but I don't think this lab was so much about taking the data as it was about analyzing it. The skills that were developed were along the lines of utilizing MATLAB to generate useful graphs and using numerical error techniques to quantitatively find error. If I were to repeat this experiment like to develop some MATLAB code that the data could be piped right in and it would generate graphs and error as the experiment ran. That'd really be a definite, dynamic improvement.