358 10 C 5
The development of Ipsix: A psychometric test for the selection and development of high quality examiners
C Dewberry, A Davies Muir, S Newell (Royal College of Paediatrics and Child Health, 5-11 Theobalds Road, London WC1X 8SH, United Kingdom)
Background: The purpose of this paper is to explain the development and validation of a psychometric test designed to predict examiner quality and to provide examiners with feedback about their
characteristics as assessors.
Summary of work: The development of Ipsix draws on theories and research on the process by which people assess and calibrate the performance of others. The validation of the instrument involved two stages. First, Rasch modelling was used to examine the assessment characteristics of over 400 examiners collectively involved in over 25,000 assessments. Second, subsets of these examiners completed Ipsix, and the performance of this test in predicting key examiner characteristics was assessed.
Summary of results: Multiple regression indicates significant associations between Ipsix scales and three indices of examiner quality.
Conclusions: Ipsix provides a novel and effective way of predicting examiner quality with respect to stringency/leniency, examiner consistency, and use of evaluative scales.
Take home messages: The quality of an examiner depends on their expertise in the field and also on their ability and style as an assessor. Ipsix is the first validated psychometric test to predict examiner quality. It can be used in both the selection and the development of high quality examiners.
10 C 6
C H Tsai, C Y Chan, W Chen, S C Liao (China Medical University Hospital, 91, Shueh-Shih Road, 404 Taichung, Taiwan)
Background: In 2007, China Medical University decided to extend the clinical education in the medical program from two to three years. To strengthen clinical education and teacher qualification, we
eachers. Summary of work: One hundres Sixty-six clerks answered a survey in March, 2008 and 159 (95.7%) of the filled surveyed were effective and analyzed.
Summary of results: Based on results of statistical analyses, we found that characteristics that described the best clinical teachers included: passionate about teaching, good patient-physician relationship, professionalism, good teaching attitude, rich professional knowledge, care for student, and effective teaching skills. Characteristics that described unpopular clinical teachers included: bad teaching attitude, not passionate about teaching, poor patient-physician relationship, and ineffective teaching skills. Conclusions: Although teaching attitude and professional knowledge were crucial, in the scenario of
-physician relationship and professionalism.
359
Take home message: A good teaching attitude and professionalism were the two qualifications for clinical teachers. Further research is called upon to explore how to help clinical teachers to develop these qualifications.
10 D Oral Presentations Other Outcomes 3 10 D 1
Using the UCLA Script-Concordance Test of cross-cultural care at the University of Michigan
ML Lypson, JA Purkiss, PT Ross, CM Grum (University of Michigan Medical School and the David Geffen School of Medicine, University of California, Los Angeles, Office of GME 2600 Green Road #150, Ann Arbor; Center for Education Development and Research UCLA; Ann Arbor, MI 48105, United States) Background: Assessment of medical student ability in cross-cultural care is mandatory given rapidly changing demographics of the North American population. External validity of a Script Concordance Test (SCT) was explored through comparison with the diagnostic pattern recognition exam (DPRE) used in the Internal Medicine clerkship, as well as other test scores.
Summary of work: We correlated results for 164 students who completed both the online SCT, and the DPRE test. SCT scores were also correlated with USMLE Step 1 and OSCE communication skills.
Summary of results: SCT mean score = 87.46% (SD=11.27, N=167), and DPRE mean = 178.09 (SD=8.98, N=164). Correlation between SCT and PR was statistically significant, though weak: Pearson's: R=0.161, p=0.040, N=164. SCT did not correlate with USMLE Step 1 (R=0.002, p=0.977) or OSCE communication skills (R=-0.092, p=0.240).
Conclusions: Given these weak or non-existent correlations, SCT may require refinement, or it may be assessing a unique clinical reasoning domain.
Take home messages: Although we found evidence of minimal overlap between reasoning assessed by SCT and that assessed by DPRE, it is also possible that issues related to cross-cultural care allow for a new and unique type of pattern recognition for disease presentation. Further assessment and validation of the SCT is warranted.
10 D 2
Conflict of interest in medical education: Where do we currently stand in residency programs? Kathryn Ross, William Iobst, Brian Hess (American Board of Internal Medicine, 510 Walnut Street, Suite 1700, Philadelphia 19106, United States)
Background: The American Board of Internal Medicine Foundation was requested by the American Board of Internal Medicine Conflict of Interest Taskforce to explore educational venues to enhance the knowledge, skills, and attitudes with conflict of interests (COI) of medical students, residents and
