Voluntary vs. compulsory student evaluation of clerkships: effect on validity and potential bias

BMC Medical Education

Table 5 Effect of authority-induced bias on response pattern of students in the compulsory group^a

Characteristics	Voluntary (N = 192 responses)	Compulsory – Non Biased (N = 29 responses)	Compulsory – Biased (N = 14 responses)	P
Overall rating^b [Mean (SD)]
Internal Medicine	53.77(7.92)	59.21(6.00)	58.96(5.85)	0.23
Surgery	49.70(7.93)	54.04(4.91)	40.71(4.95)	0.10
Pediatrics	55.83(7.90)	59.21(4.36)	49.20(7.83)	0.36
Ob-Gyn	56.60(6.59)	55.60(6.26)	49.83(15.15)	0.32
Primary Care	55.33(8.88)	57.87(5.51)	51.21(−--)^c	0.79
Neurology	49.86(9.92)	52.46(9.95)	46.20(−--)^c	0.82
Psychiatry	52.33(8.92)	58.71(6.40)	–	0.19
All Clerkships	53.19(8.69)	56.38(6.17)	50.88(10.09)	0.09
Presence of Comments^d [N (%)]	44(22.9%)	5(17.2%)	1(7.1%)	0.32
Cronbach alpha	0.945	0.946	0.944	NA

Abbreviations: Ob-Gyn Obstetrics and Gynecology, NA not applicable
^aComparisons involved only the last evaluation administration (including the sham question) in the compulsory group
^bOverall rating per clerkship was calculated for respondents in the voluntary group, where the number of obtained evaluations per clerkship is less than 49 (the total number of students in the class) depending on the response rate per clerkship
^cNo standard deviation is computed because only one student completed the evaluation for Primary Care
^dComments in the voluntary group were examined on the whole number of obtained responses (192)

ISSN: 1472-6920