Quantitative analysis of a Māori and Pacific admission process on first-year health study

BMC Medical Education

Table 8 Multiple regression analysis results – linear regression^a

Multivariate analysis results	First year tertiary students		First year bachelor students
Multivariate analysis results	2009 – 2012 (n = 368)		2009 – 2012 (n = 242)
	Mean estimate (95 % CI)	P value	Mean estimate (95 % CI)	P value
GPA Eight Courses
NCEA Rank Score (per 20 point increase)	0.26 (0.18, 0.34)	<0.0001	0.40 (0.30, 0.50)	<0.0001
Followed MAPAS advice
No	0.00		0.00
Yes	1.17 (0.57, 1.78)	0.0002	1.09 (0.45, 1.73)	0.0009
Any 2 sciences
No	0.00		0.00
Yes	0.65 (0.15, 1.15)	0.0116	0.39 (−0.29, 1.08)	0.2603
MAPAS Maths test (per 10 % increase)	0.14 (0.02, 0.26)	0.0186	0.08 (−0.07, 0.22)	0.2885
GPA Core 4 Courses
NCEA Rank Score (per 20 point increase)	-	-	0.38 (0.26, 0.50)	<0.0001
Followed MAPAS advice	-	-
No			0.00
Yes	-	-	1.14 (0.60, 2.04)	0.0004
Any 2 sciences	-	-
No			0.00
Yes	-	-	0.64 (−0.13, 1.41)	0.1027
MAPAS Maths test (per 10 % increase)	-	-	0.15 (−0.02, 0.31)	0.0765

^a Adjusted for MAPAS interview year, gender, ancestry and school decile. For GPA (a continuous outcome variable), its mean change associated with the change in a linear predictor was estimated with 95 % confidence interval. For a continuous predictor variable, this gave the difference in means with either 20 point (NCEA Rank Score) or 10 % (MAPAS Maths percentage mark) increase in the predictor. For a categorical predictor, this gave the difference in means between the current and reference categories (i.e. yes vs. no). The null hypothesis was that there was no change in the mean response (i.e. Δ = 0)

ISSN: 1472-6920