Skip to main content
. 2017 Aug 27;7(8):e017276. doi: 10.1136/bmjopen-2017-017276

Table 3B.

Linear regression results for Year 2–3 GPA (2002–2012 cohort) academic outcome for MAPAS and GENERAL subgroups

Model Predictor variable (ref) Comparison Year 2–3 GPA (2002– 2012 cohort)
MAPAS (n=315)
Year 2–3 GPA (2002–2012 cohort)
GENERAL (n=1167)
Overall p-Value Mean difference 95% CI Overall p-value Mean difference 95% CI
1 Ethnicity
(reference group Māori for
MAPAS, Pākehā/European for
GENERAL)
Māori <0.0001 −0.616 −1.201 to to 0.031
Pacific 0.0109 −0.425 −1.018 to 0.168 −2.226 −2.939 to to 1.513
Asian −0.191 −0.358 to to 0.025
Other/missing −0.524 −0.852 to to 0.196
5 n=302 n=1104
School decile (high 8–10) Medium (4–7)
Low (1–3)
Auckland school (yes) No
Type of admission (SL) AA
Bridging programme (no) Yes <0.0001 −0.927 −1.209 to –0.654 0.0036 −1.083 −1.812 to to 0.355
Entry pathway (undergraduate) Graduate <0.0001 0.577 0.377 to 0.777
Admission GPA/GPE Per point increase <0.0001 0.754 0.647 to 0.861 <0.0001 0.977 0.891 to 1.063

Logistic regression model has controlled for year of admission, gender and age at admission. Predefined predictors were added to the baseline model in sequential order to estimate their joint independent effects on the outcome. Each model was initially run with all the specified predictors, then rerun with stepwise selection to include significant predictors only and obtain final estimates of effect size. Model-adjusted estimates of ORs (compared with the reference level), 95% CI and associated individual p-values (in symbols) were reported.

AA, alternative admission (ie, non-school leaver); GPA, Grade Point Average; GPE, Grade Point Equivalent; MAPAS, Māori and Pacific Admission Scheme; SL, school leaver.