Indigenous Students’ Access to Post-Secondary Education in British Columbia - Evidence from BC PEN Linked Data

The contribution of this report to the existing literature is, therefore, to leverage the power of the BC PEN data to track Indigenous students’ rates of access to PSE over time and to advance our understanding of how PSE access rates are related to the rich set of measures available in the BC PEN data, including student, neighbourhood, K-12 program, school, and skill assessment characteristics, a. [...] The report then presents the descriptive findings followed by the regression results in two parts, corresponding to the two sets of models estimated: access to any PSE and access to the specific PSE program of 2 Access to PSE is conventionally defined in terms of the first program entered, as here. [...] For the presentation of the findings, we use standard procedures to transform the estimated coefficients into average marginal effects, which represent the difference in the expected probability of the outcome associated with a one unit change in the variable in question, or in comparison to the omitted, or “baseline” category for the categorical variables employed. [...] The first consists of the descriptive analysis, the second presents the findings of the logit regression where the outcome is access to any PSE program, and the third presents the findings of the multinomial logit regression where the outcomes are the different kinds of PSE program first accessed. [...] Similar to the logit models previously seen, the results shown represent the transformation of the estimated coefficients into average marginal effects, or the average difference in the probability of accessing that type of program associated with a one unit change in the explanatory variable (as compared to the omitted reference category for the categorical variables), along with the estimated st.