In sample surveys, we randomly select sampling units from a population and estimate the parameter, which is the characteristic of interest. However, nonresponse will occur due to various reasons, those data cannot be a representative of the population, and it can have a significant effect on the survey results. Thus, we need a solution to nonresponse.In this study, we develop the nonresponse imputation class for item nonresponses in a sample survey called Time Use Survey of Students, impute them by using two hot-deck imputation methods, and then compare the effectiveness of the results from those methods.