Skip to main content
. 2015 Jul 15;30(9):1307–1312. doi: 10.1007/s11606-015-3337-z

Table 3.

Estimates for the Final Multilevel Model Predicting Internal Medicine Subject Examination Scores

Coefficient Standard error p value Interpretation
Grand mean 77.44 0.16 < 0.01 Grand mean for internal medicine subject examination scores
Internal medicine rotation length 0.23 0.08 < 0.01 Each additional week in internal medicine increases predicted subject examination performance by 0.23 points
Average Step 1 score 0.24 0.03 < 0.01 Each 1-point increase in a school’s mean Step 1 performance increases predicted subject examination performance by 0.24 points
Completed surgery rotation 1.58 0.10 < 0.01 Students who completed surgery do better by 1.58 points
Completed pediatrics rotation 0.91 0.13 < 0.01 Students who completed pediatrics do better by 0.91 points
Completed family medicine rotation 0.87 0.11 < 0.01 Students who completed family medicine do better by 0.87 points
Completed obstetrics and gynecology rotation 0.14 0.12 0.26 Students who completed obstetrics and gynecology do better by 0.14 points (not statistically significant)
Completed psychiatry rotation 0.12 0.10 0.26 Students who completed psychiatry do better by 0.12 points (not statistically ignificant)
Step 1 score 0.25 0.002 < 0.01 Each 10-point increase in Step 1 scores is associated with a 2.5-point increase in predicted subject examination performance

Italics indicate school-level variables predicting the grand mean for internal medicine subject examination scores