Analysis methods for results in gamma-ray astronomy.

The procedures in current use for analyzing the results of gamma-ray astronomy experiments are assessed. Two formulas are proposed for estimating the significance of positive observations in searching gamma-ray sources or lines. The correctness of the formulas is tested with the aid of Monte Carlo simulations. One formula is derived by immediately estimating the standard deviation of the observed signal N_S; the other is derived by applying the method of the statistical hypotheses test.