The previous studies have shown that most of the results of donation were not significant except they were built on the premise of the veil of ignorance. In this study, we modify the experimental design to exclude the veil of ignorance which is the factor of risk. Furthermore, we consider that the target of donation people care about is not limited to money but the primary social goods. By conducting a lab experiment of the dictator game, we try to investigate whether people make donations is because of the needs of others. The experiments are divided into the control and the experimental group. The design of the control group is the same as the traditional dictator game, while the experimental group is to establish a threshold value which decides whether or not people can obtain the primary social goods. The results of experiments support the theoretical predictions. Subjects in the control group show egoism as in the general experiments of the dictator game. But the subjects in the experimental group show that people really care about the needs of others, and are willing to sacrifice a bit of their payoffs to help people in need.