This paper examines the reliability of the knowledge measurement methods about radiation and also analyzes the factors which affect people's anxiety. A true-false test was used to measure people's knowledge of radiation since it is considered to be a more objective method than self-report method. In this paper two types of true-false test were made and their accuracy rates were compared. Each test contained question items that evaluate scientific knowledge of radiation but were worded differently. As a result of this experiment, some question item pairs which evaluate the same knowledge but had different wordings showed a significantly different accuracy rate, which means that the accuracy of the true-false test is relative to the wording of the questions. As a solution to this problem, item response theory was used to reduce the influence of the wording on this knowledge measurement test. Finally, the relationship among anxiety and other variables was analyzed. In this study interest in social problems didn't have a significant effect on anxiety. However, activeness of information gathering and risk perceptions had a positive effect. It was also determined that hearing from specialists and having an interest in science or knowledge of radiation reduced anxiety.