Mathers Jonathan, Sitch Alice, Parry Jayne
Institute of Applied Health Research, College of Medical and Dental Sciences, University of Birmingham, Edgbaston, UK.
Med Educ. 2016 Oct;50(10):1033-44. doi: 10.1111/medu.13082.
Medical schools are increasingly using novel tools to select applicants. The UK Clinical Aptitude Test (UKCAT) is one such tool and measures mental abilities, attitudes and professional behaviour conducive to being a doctor using constructs likely to be less affected by socio-demographic factors than traditional measures of potential. Universities are free to use UKCAT as they see fit but three broad modalities have been observed: 'borderline', 'factor' and 'threshold'. This paper aims to provide the first longitudinal analyses assessing the impact of the different uses of UKCAT on making offers to applicants with different socio-demographic characteristics.
Multilevel regression was used to model the outcome of applications to UK medical schools during the period 2004-2011 (data obtained from UCAS), adjusted for sex, ethnicity, schooling, parental occupation, educational attainment, year of application and UKCAT use (borderline, factor and threshold).
The three ways of using the UKCAT did not differ in their impact on making the selection process more equitable, other than a marked reversal for female advantage when applied in a 'threshold' manner. Our attempt to model the longitudinal impact of the use of the UKCAT in its threshold format found again the reversal of female advantage, but did not demonstrate similar statistically significant reductions of the advantages associated with White ethnicity, higher social class and selective schooling.
Our findings demonstrate attenuation of the advantage of being female but no changes in admission rates based on White ethnicity, higher social class and selective schooling. In view of this, the utility of the UKCAT as a means to widen access to medical schools among non-White and less advantaged applicants remains unproven.
医学院校越来越多地使用新颖工具来筛选申请者。英国临床能力倾向测试（UKCAT）就是这样一种工具，它通过一些可能比传统潜在能力衡量指标受社会人口统计学因素影响更小的构念来测量有助于成为一名医生的心理能力、态度和职业行为。各大学可根据自身情况自由使用UKCAT，但已观察到三种大致模式：“临界值”模式、“因素”模式和“阈值”模式。本文旨在进行首次纵向分析，评估UKCAT的不同使用方式对向具有不同社会人口特征的申请者发出录取通知的影响。
采用多水平回归模型分析2004年至2011年期间申请英国医学院校的结果（数据来自大学和学院招生服务中心），并对性别、种族、学校教育、父母职业、教育程度、申请年份和UKCAT使用方式（临界值模式、因素模式和阈值模式）进行了调整。
除了以“阈值”方式应用时女性优势出现明显逆转外，UKCAT的三种使用方式在使选拔过程更公平方面的影响并无差异。我们尝试对UKCAT阈值模式使用的纵向影响进行建模，再次发现了女性优势的逆转，但未显示出与白人种族、较高社会阶层和选择性学校教育相关的优势有类似的统计学显著降低。
我们的研究结果表明女性优势有所减弱，但基于白人种族、较高社会阶层和选择性学校教育的录取率没有变化。鉴于此，UKCAT作为一种扩大非白人及弱势群体进入医学院校机会的手段，其效用仍未得到证实。
