First, in exploratory factor analysis, we conducted 25 items to clarify the factor structure of questions related to radiation exposure. In the procedure for exploratory factor analysis, the number of factors was estimated using the eigenvalues and the decay situation of the scree plot, parallel analysis, and MAP, and the four-factor structure that was most supported by each estimation method was adopted. Next, factor extraction was performed using the maximum likelihood method with Promax rotation. The criterion for factor loadings was set at 0.35, and items that did not contribute to any factor and items that contributed to multiple factors were excluded (6 items; Table1).

Factor 1: Anxiety about radiation exposure. This factor showed high loadings for items such as, “Anxiety about radiation exposure sometimes makes it difficult to concentrate on observation and learning during practical training” and “After practical training, I sometimes feel anxious at home or before going to bed, fearing possible radiation exposure”.

Factor 2: Interest in radiation exposure. This factor showed high loadings for items such as, “I am not very interested in the amount of radiation I will be exposed to during my practical training,” and “I would like to be informed of my own radiation exposure during training only when it exceeds the standard”.

Factor 3: Knowledge about radiation exposure. This factor showed high loadings for items such as, “I believe I have accurate knowledge about radiation and exposure in the medical field” and “I can accurately judge whether or not to perform tests involving radiation exposure”.

Factor 4: Management of radiation exposure. This factor showed high loadings for items such as, “I would like my radiation exposure during practical training to be managed using quantitative, definitive values” and “I believe that the use of radiation in medicine is highly effective”.

The results of the mixed-design analysis of variance examining within subjects the effect of the lecture revealed significant main effects for survey period (pre vs. post) (F(1, 89) = 7.14, p < .0001, η2p = .07) and factor (anxiety about radiation exposure, interest, knowledge, management) (F(3, 267) = 59.48, p < .0001, η2p = .43). A significant interaction between survey period and factor was also observed (F(3, 267) = 2.27, p < .0001, η2p = .03). Following up on the significant interaction with simple main effect tests for survey period within each factor, the results indicated that scores for “knowledge about radiation exposure” and “management of radiation exposure” were significantly higher in the post-survey than in the pre-survey (p < .001 for both, Table 2). Conversely, no significant change was observed in “interest in radiation exposure” and “anxiety about radiation exposure” between the pre- and post-surveys (Table 2).

Furthermore, the maximum amount of radiation exposure allowed differs significantly between genders, with exposure to radiation for female students being strictly managed. To verify whether these factors also affect the students’ practical training, we conducted a mixed-design analysis of variance with 2 (gender: male or female) × 4 (factors: anxiety, interest, knowledge, and management of radiation exposure) using the difference of scores between post and pre (post–pre scores). Only the main effect of “factors” was significant (F(3, 267)=5.46, p < .0001, η2p = .10), and there was no interaction effect (Table 3).

In addition, since we used the difference of scores between post and pre, we tested whether the scores were significantly greater than zero for factors using the adjusted Bonferroni method (p < .05). As results, after the lecture, students showed significant increases in the scores for the other factors “interest in radiation exposure (t(89)= 2.07, p = .0205)”, “knowledge about radiation exposure (t(89)= 7.94, p < .0001)”, and “management of radiation exposure(t(89)= 5.31, p < .0001)”. (Figure 1).

Differences (post–pre) in anxiety, interest, knowledge, and management of radiation exposure between the gender of students. Scores for anxiety decreased and scores for interest, knowledge, and management increased after the lecture for both males and females. Error bar indicates MSE (mean squared error)

The mixed-design analysis of variance investigating the effect of clinical training selection using the difference in scores between post and pre (post–pre) revealed a significant interaction between training selection and factor (F(3, 267) = 3.36, p = .0309, η2p = .02). Simple main effect tests showed a significant main effect of training selection for “knowledge about radiation exposure (F(1, 185) = 4.39, p = .0038, η2p = .02)” and “management of radiation exposure (F(1, 185) = 6.34, p = .0013, η2p = .03)” (Table 4). In addition, we tested whether the difference in scores was more significant than zero for each of the factors using the adjusted Bonferroni method (p <.05). The results showed that the scores for knowledge and management were significantly greater than zero regardless of whether the selection of department. The score for interest was a marginally significant trend (t(89)=2.07, p = .0853) only for students who selected radiology or cardiology (Figure 2).

Differences (post–pre) in anxiety, interest, knowledge, and management of radiation exposure among students selecting radiology and cardiology vs. others. Comparing pre and post-lecture scores of interest increased for only students who selected radiology or cardiology. Radiology and cardiology n=57, others n=38. Pre = before the lecture; post = after the lecture. The asterisk (*) indicates a significant difference (p < .05). Error bar indicates MSE (mean squared error)