Ethno-linguistic patterns of degree completion in BC universities R. Sweet, A. Pullman, M. Adamuti-Trache, & K. Robson Table 1. Regression estimates of BA/BSc completion, AME Baseline Multilevel Logit Multilevel Logit Logit Without IP weights With IP Weights Model 0 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 0.065*** -0.012 0.0049 -0.0029 -0.026 -0.0100 -0.017 (0.017) (0.012) (0.013) (0.016) (0.015) (0.016) (0.017) -0.0042 0.019 0.041* 0.023 0.0072 0.034 0.017 (0.017) (0.015) (0.019) (0.027) (0.015) (0.023) (0.030) 0.19*** 0.016 0.043* 0.027 -0.0094 0.028 0.015 (0.009) (0.016) (0.019) (0.015) (0.018) (0.017) (0.012) -0.020 -0.20*** -0.18*** -0.15*** -0.25*** -0.21*** -0.19*** (0.022) (0.031) (0.024) (0.031) (0.044) (0.036) (0.045) -0.022 -.073*** 0.068*** 0.083*** -0.095** -0.088** -0.100** (0.027) (0.015) (0.021) (0.015) (0.029) (0.031) (0.032) 0.049** -0.022 0.016 0.0046 -0.037* 0.011 0.0030 (0.019) (0.013) (0.011) (0.012) (0.018) (0.018) (0.012) -0.17* -0.18* -0.12 -0.12 (0.081) (0.079) (0.074) (0.073) -0.27*** -0.27*** -0.20* -0.19** (0.070) (0.061) (0.078) (0.070) 0.13*** 0.11*** 0.15*** 0.14*** (0.011) (0.015) (0.013) (0.018) 0.23*** 0.21*** 0.26*** 0.24*** (0.022) (0.027) (0.023) (0.029) 0.28*** 0.26*** 0.31*** 0.29*** (0.029) (0.034) (0.027) (0.035) 0.037*** 0.031* 0.0027 0.00014 (0.010) (0.015) (0.012) (0.017) 0.084*** 0.085*** 0.050* 0.052 (0.022) (0.025) (0.024) (0.028) 0.13*** 0.13*** 0.10*** 0.10** (0.029) (0.034) (0.029) (0.036) 0.070** 0.067** 0.038 0.039 (0.021) (0.023) (0.023) (0.026) Ethnolinguistic groups (ref: English) European South Asian Chinese Korean Tagalog/Vietnamese Other Institutional type at entry (ref: RIU) TIU CT High-school GPA (ref: 1-24 per.) 25-49 percentile 50-74 percentile 75-99 percentile Interaction terms (Ref: RIU) TIU*25-50 TIU*50-75 TIU*75-100 CT*25-50 Canadian Journal of Higher Education | Revue canadienne d’enseignement supérieur 49:3 (2019) Ethno-linguistic patterns of degree completion in BC universities R. Sweet, A. Pullman, M. Adamuti-Trache, & K. Robson Baseline Multilevel Logit Multilevel Logit Logit Without IP weights With IP Weights Model 0 Controls Observations Model 3 0.037 -0.0008 0.0061 (0.041) (0.040) (0.044) 0.071* 0.075* 0.043 0.047 (0.031) CT*75-100 Model 2 (0.038) CT*50-75 Model 1 Model 2 Model 3 0.033 (0.036) (0.032) (0.037) Included Model 1 Included 16,795 Standard errors in parentheses; * p < 0.05, ** p < 0.01, *** p < 0.001; The baseline logit model presents the results without accounting for the clustering effect of institution of entry (n=22). The multilevel logit models demonstrate how the results change with the inclusion of the main independent (Model 2) and control variables (Model 3). Model three controls for: gender; Indigenous ancestry; prior K-12 ESL training; if a student’s high school was located in Metro Vancouver; timing of entry to the BC K-12 system; field of study at PSE entry, and timing of PSE entry. Canadian Journal of Higher Education | Revue canadienne d’enseignement supérieur 49:3 (2019)