Measures of radioactivity: a tool for understanding statistical data analysis

Phys. Department, University of Siena Experimental setup

The measures are realized by using a commercial Geiger counter usually used for dosimetry. It can detect alpha, beta and gamma rays and it is possible to collect data without any unit conversion (counts) and download them into a computer. The detector and the sources are aligned by means of supports on an optical bench. The sources are shown with their supports (from left): a) small sheets of Uranium in plastic holder m1 = 0.152 g b) sheets U in plastic holder m2 = 0.888 g c) fluorescent marble

Vera Montalbano

An early experience
The starting point was the Pigelletos Summer School of Physics [1-4] . Forty students from high school are selected to attend a full immersion summer school of physics in the Pigelleto Natural Reserve, on the south east side of Mount Amiata in the province of Siena. In the 2009 school, titled The Achievements of Modern Physics, a learning path on radioactivity was proposed to two small groups of students. After a brief introduction to the nuclear phenomenology, they were involved in measures of radioactivity from a weak source of U in order to characterize the emitted ionizing radiation by using a educational device provided by a school. A teacher was impressed by students involvement and suggested of elaborating a learning path on Nuclear Physics in which the students are active and perform directly measures of radioactivity.

Our target
In order to integrate this learning path in the ordinary school program, the activities are planned for the last year of high school (5a Liceo Tecnologico) after the lessons on electromagnetism. Teachers make few lessons in class in which present and discuss the phenomenology of nuclear physics. Afterward, the experimental setup is presented in order to perform measure of background radioactivity in the laboratory. The next step is to divide students in small groups that are involved in measures of , and emission from weak sources of U and a very weak source with trace of uranium ore. During measurements, it rises soon the necessity of performing many measures and the problem of valuating uncertainties. The students are ready to recall all their knowledge on statistics for facing this problem.

A brief introduction to nuclear phenomena

The lessons give an historical sight on the discovery of radioactivity, and connect the previous knowledge of students to these phenomena so far from their daily experience. The main topics are the following: - Discovery of radioactivity - Pierre e Marie Curie - Neutron discovery and atom structure - Dimensions of atom and its constituents - Natural radioactivity phenomenology (in particular , and rays) - Isotopes - Forces into atoms and nuclei (how strong or weak with numerical examples) - Nuclei can die - Law of radioactive decay - Natural radioactive chains - Binding energy and mass of a bound state - Equivalence mass-energy - Binding energy per nucleon - Fission and related topics - Fusion and related topics Related topics can be very stimulating for students, specially in these times for our society

An introduction to statistical data analysis

interessante stimare qual la varianza della media m

Prime misure in classe

Fondo =18.08
d1=d0 m1 = 0.152 g m2= 0.888 g 2753 37 2485 41 326 5970 54 4925 143 2039

0.05

N=25
d3=d0+26.4 244 286 215 714 608 227

d1=d0+4.25 24821 2108 213 184855 157134 685

Let us start from a question: There are limits to the precision of a measurement? By answering to this question, we can fix the cases in which the use of statistical data analysis make sense. The first step is to recall all previous knowledge of students on statistics (there are many but almost never used in physics by teachers), such as [5,6] - Distributions - Representation of casual phenomena - Mean, median, variance and standard deviation - Bernoulli, Poisson and Gauss distributions and examples At this point, it is possible to introduce the pupils to some elements of sampling theory, such as - The existence of a parent distribution - What is a good statistical sample - Mean and variance of a sample Now we are ready to deal with the statistical data analysis of our measures. The data are collected in the meantime by the Geiger detector and can be downloaded in a computer. The teacher can divide the data in sets in order of giving different tasks to any group of students. Some examples: - testing samples by giving 10, 20, 30, 50 data from the background and compare means and standard deviations of these samples with the mean and standard deviation of much more measures (e.g. more then 1000) - Perform measures for short period in order to verify that radioactive decay follows Poissons distributions - Verify that for mean values greater than the Poissons distribution become indistinguishable from the normal one In particular, can be very interesting to show that in case of very weak source it is possible to quantify differences from background smaller than the uncertainties related to the physics of radioactive decay by performing many measurements and by using statistical data analysis.

Si pu far vedere che

2 sm n 1 ( x m) 2 n( n 1) i 1

Quindi nel caso in cui si hanno misure i cui risultati fluttuano, una volta verificato che non ci sono errori sistematici , il risultato della misura si scrive:

m sm

References
[1] [2] [3] [4] [5] [6] A. Porri, R. Benedetti, E. Mariotti, V. Millucci, V. Montalbano, Esperienze significative del Progetto Lauree Scientifiche di Siena, Proceeding ComunicareFisica, Trieste 1 6 ottobre 2007 A. Porri, R. Benedetti, E. Mariotti, V. Millucci, V. Montalbano, La Scuola estiva del Pigelleto e I Giocattoli al Liceo, in atti del XLVI National Congress AIF, Suppl. n. 3 La Fisica nella Scuola , 130, 2008 R. Benedetti, S. Di Renzone, E. Mariotti, V. Montalbano, A. Porri, Un percorso di orientamento sulla fisica moderna, Comunicare Fisica 2010, in corso di stampa R. Benedetti, E. Mariotti, V. Montalbano, A. Porri, Active and cooperative learning paths in the Pigelletos Summer School of Physics, Twelfth International Symposium Frontiers of Fondamental Physics (nearby poster ID 91) P. R. Bevington Data Reduction and Errors Analysis for Physical Science, McGraw-Hill Book Company, New York 1969 http://stattrek.com/Lesson3/SamplingTheory.aspx?Tutorial=Stat