Summary: In this paper, the fundamental principle of double factor variance analysis is briefly introduced, and it is pointed out that an experiment can be affected by various factors, the same factor can influence the results at different levels. A new concept of interactions is presented. Two factors non-repetitive experiments with and without interactions are studied. Excel is used for the experimental designs according to the specific situations. Experimental data of electronic utilities in different districts and traffic times and road sections are analyzed by variance analysis. Hypothetic tests and statistical inference are given for the parameters of a regression model, and judgments on different factors with different levels are obtained.