The questionnaire has been quite common in life, and building a scientific quan-titative criteria for analyzing the questionnaire is very important. At present, forsome quantitative criteria, but it only shows how to analyze the data simply, whileignoring the hidden value in the absence of some important information.In this paper, aiming at the problem of missing data in questionnaire, we pro-pose a new method for mending missing data, which can make the correlation matrixof the subject data and the population data maintain consistence, make the new datamore accurate, and have more self-space than the conventional methods. By usingthe independence test of contingency table in the questionnaire data, we find thatgender plays an important role in the recognition of work ability. Bared on the logis-tic model and SAS program, we simplify the questionnaire, and keep the integrity oforiginal data. We finally conclude that the first task of honest work is the people'slivelihood, and the results of the paper can provides a reliable basis for the later setof questionnaire. |