Generalizability of a Composite Student Selection Procedure at a University-Based Chiropractic Program

Abstract

Good generalizability for a composite admission to a chiropractic program was found. Optimal weighting and adequate sampling are important for obtaining optimal generalizability. Limitations and suggestions for future research are discussed.

Assessments of Non-academic Attributes in Applicants for Undergraduate Medical Education: an Overview of Advantages and Limitations

“…MMIs have an alpha of 0.61-0.98; inter-rater correlations of 0.41-0.91; test-retest correlations of 0.34-0.72; and generalizability over other settings (G coefficient) of 0.51-0.88 [10,[22][23][24]. SJTs have an alpha of 0.43-0.94 and test-retest of 0.20-0.92 [25,26], and SCs have an alpha of 0.88-0.91; inter-rater correlations of 0.62-0.77; test-retest correlations of 0.59-0.43; and G coefficient of 0.82 [13,27]. Finally, scores on MMIs, SJTs, SCs, and personality traits have been reported to correlate with various aspects of later clinical performance (see Table 1 for an overview of published reviews and more recent studies).…”

“…SJTs are difficult to design, and their scoring is complicated because there are no "correct" answers [12]. SCs combine multiple-choice question testing, interviews, writing essays, carrying out MMI-like assignments, listening to a lecture, and answering questions about its content [13]. Tests of personality traits usually focus on conscientiousness (responsibility, self-discipline) [14,15], emotional intelligence (monitoring one's own and others' emotions) [16][17][18] and empathy (understanding and responding to another person's concerns) [19].…”

Testbaseret optag til videregående uddannelser

“…Resultatet viste så høj grad af uoverensstemmelse i bedømmelsen af ansøgningerne (foretaget af to bedømmere), at resultatet for generaliserbarhed (G = 0,10) lå vaesentligt under den acceptable graense. Til sammenligning blev der fundet god generaliserbarhed ved anvendelsen af strukturerede samtaler og for kombinationen af vurdering af kvalifikationer, samtale og paratvidentest (G = 0,82) (O'Neill et al, 2009). Derudover kan det vaere en saerdeles administrativ tung og ressourcekraevende opgave at laese motiverede ansøgninger igennem, isaer hvis der skal vaere mindst to bedømmere per ansøgning.…”

Fuldt nummer

“…Resultatet viste så høj grad af uoverensstemmelse i bedømmelsen af ansøgningerne (foretaget af to bedømmere), at resultatet for generaliserbarhed (G = 0,10) lå vaesentligt under den acceptable graense. Til sammenligning blev der fundet god generaliserbarhed ved anvendelsen af strukturerede samtaler og for kombinationen af vurdering af kvalifikationer, samtale og paratvidentest (G = 0,82) (O'Neill et al, 2009). Derudover kan det vaere en saerdeles administrativ tung og ressourcekraevende opgave at laese motiverede ansøgninger igennem, isaer hvis der skal vaere mindst to bedømmere per ansøgning.…”