68-2

43/110

(self-report)1~ 25％（n=60）25～50％（136）50～75％（159）75～100％ (169)(p=0.04).4 months inadvance～10(15.2%)23(16.9%)29(18.2%)25(14.8%)3 monthsin advances90 (54.5%)41 (43.6%)No. of studentsmore than 4 months 3 months inadvance～9(13.6%)23(16.9%)31(19.5%)50(29.6%)2 months inadvance～6(9.1%)20(14.7%)27(17.0%)39(23.1%)Time of starting preparationin advances92 (53.8%)55 (55.0%)171100a month inadvance～3(4.6%)6(4.4%)16(10.1%)25(14.8%)3 monthsin advances79 (46.2%)45 (45.0%)for CBT, it was found that the time was earlier among students with favorable results and later among those with poor results. Among 196 students realizing that they were upper lanked, 135 (68.9%) had started their preparation for CBT 4 months in advance (Table 5: light gray area). In contrast, among 328 realizing that they had a low rank, 188 (57.3%) had started it 3 months in advance (Table 5: dark gray area). Especially, 14.8% of all students realizing that they were among the bottom quartile had started it 1 month in advance, which was after the beginning of the summer vacation (Table 5). On comparing the time of starting preparation in each science-elective-based group, the rate of students starting it more than 4 months in advance did not vary between without and with COVID-19 in the biology group, whereas there was an increase of 8.3% with COVID-19 in the physics group (Table 6).4. The ways to spend the summer vacationIn the with COVID-19 group, there were no students who had participated in the summer camp, whereas 170 of the without COVID-19 group had participated in it, and their IRT scores were significantly lower than those of the former (p=0.03).Furthermore, in the without COVID-19 group, the IRT score tended to be low when the sum of the days spent participating in the summer camp, working part time, and travelling was 10 or more Grades in the third year■ Students with upper lank who started 4 months in advance　 ■ Students with lower lank who started 3 months in advanceNo. of studentsmore than 4 months Physics Group Biology Group16594Table 5　Relationship between performance in the third year and the time of starting preparation for CBTTable 6　Time of starting preparation for CBT in each science-elective-based groupWith COVID-19More than 6 months 5 months inadvance～16(24.2%)47(34.6%)45(28.3%)20(11.8%)in advance～22(33.3%)17(12.5%)11(6.9%)10(5.9%)Without COVID-19Time of starting preparationin advances75 (45.5%)53 (56.4%)The IRT scores of students who had not taken any practice examination were significantly lower in both the without and COVID-19 groups (p<0.01, p=0.01), and the rate of having a poor CBT score was significantly higher among these students (p<0.01) (Table 7). 5. Multivariate analysis to identify risk factors associated with poor CBT scoresWith , , , , , and < prac-tice examination (taking vs. not taking)> that revealed significant differences in univariate anal-ysis as covariances, multivariate analysis was performed to identify risk factors associated with poor CBT scores. Then, , , and < practice examination (4.58)> were identified as such factors (Table 8). Thus, it was shown that students with poor performance in the third year, starting preparing for CBT late, and not taking any practice examination are more likely to have poor CBT scores. 135