The study population comprised all standard entry students from 1985 to 2011 (students directly from high school with no prior tertiary study) (N = 2839). They were divided into 2 cohorts-those prior to 1999 who were selected on the basis of academic performance alone (N = 1402) and those after 1999 selected on the basis of a combination of academic performance, the score from a highly structured interview and the score on an aptitude test-the UMAT (N = 1437) [1
]. Academic performance before 1998 was assessed by a Tertiary Entry Score (TES) calculated from the best of 3 to 5 subjects, with a maximum score of 510. From 1999 onwards it was assessed by Tertiary Entrance Rank (TER) (maximum score 99.95) which is calculated on the basis of the total number in the cohort and the TES distribution for that year. All analyses were conducted with the inclusion of those students selected through the special entry quota pathways which comprised 40 students prior to 1999 (2.9%) and 276 students from 1999 onwards (19.2%).
The basic format and components of the highly structured interview developed at UWA have been reported previously [2
]. In 2007 a change in the interview scoring was instituted so that each of the 7 components of the interview score was ranked on a 6 point score rather than a 4 point score. The total and component interview scores have therefore been analysed with all values from 2007 standardised to a 28 point scale rather than the raw interview score. Only 2 of the 7 components have been consistently assessed since 1999-the global communication skills score and the motivation/commitment score-and so sub-analyses are also presented against the standardised score for each of these components.
Even though the total UMAT score alone was used in the ranking process, each of the three component scores, UMAT1 (Logical reasoning and problem solving), UMAT2 (Understanding people) and UMAT3 (Non-verbal reasoning) have different and independent constructs [3
] and have therefore been independently evaluated in this study together with the total score.
An Index of Community Socio-Educational Advantage (ICSEA) has now been developed for all schools in Australia [4
] and is listed for each school on the MySchool website http://www.myschool.edu.au/
. In this study, in an attempt to discern any changes in student socio-economic background pre and post introduction of the new admission process, we have imputed for all Western Australian students an ICSEA score on the basis of the high school in which they completed their TES or TER. The score is calculated from a number of variables which include parental occupation, parental school education level, parental non-school education level, percentage of families that are one parent families with dependent offspring only, percentage of occupied private dwellings with no internet connection, percentage of Aboriginal enrolments, an accessibility/remoteness index, and the percentage of students with both a language background other than English and parents with an education level of Year 9 equivalent or below. Every school has an ICSEA value on a scale which has a mean within Australia of 1000 and a standard deviation of 100. ICSEA values range from around 500 (representing extremely disadvantaged backgrounds) to about 1300 (representing schools with students from very advantaged backgrounds). The value on the scale assigned to each school is the averaged level for all students in that particular school. School of origin (and therefore ICSEA score) was unavailable for approximately 9% of the cohort.
Region of origin was determined from country of origin according to major regional groups as outlined in the Australian Standard Classification of Countries for Social Statistics [5
]. Data were available for 2821 students. Given the relatively small numbers of students in some groups they have been collapsed into 5 groups for analysis-those from Oceania (Australia, New Zealand, Papua New Guinea and proximate Pacific islands), UK and Ireland, NE and SE Asia, Southern Asia (India, Pakistan, Sri Lanka and Bangladesh) and Other.
Univariate comparisons of each demographic characteristic or each selection criteria utilised either independent sample T-tests, chi square analysis or one-way analysis of variance (with post-hoc comparisons by Bonferroni correction), as appropriate. Multivariate analyses utilised generalised linear modelling (GLM) to assess the main effects of gender, high school of origin, region of origin or special entry pathway status entered as predictive factors, ICSEA score entered as a predictive covariate and TER, TES, UMAT or interview scores (or their component parts) entered as dependent variables. All analyses were carried out utilising Predictive Analytics SoftWare Statistics Release 18.0.1, 2009.