The two scatterplots below focus on just two institutional metrics: graduation rate and net price, color-coded by Carnegie classification and funding type. (Net price is a metric that reflects the cost of tuition and fees minus the average amount of student aid received by students who receive aid, and is meant to better reflect the actual cost of attendance for undergraduates who are eligible for student aid.) You can filter institutions according to the architecture-related majors they offer, but keep in mind that these are again institution-level statistics, and that programs in architecture or related majors at any given school may have different graduation rates and net prices.
Although there is considerable variation, the overall trend in these two charts shows that institutions with a higher net price are somewhat more likely to have a higher graduation rate. This is not necessarily a causal relationship, as the net price is likely correlated with other factors, such as the socioeconomic or demographic characteristics of students who more often attend a given institution.
When we look at how this breaks down by Carnegie classification, there is considerable variation but also some clusters. In particular, the dense green cluster indicates that Associate’s level institutions consistently tend to be less expensive and have lower graduation rates than other kinds of institutions. Within each Carnegie classification, institutions with a higher net price are somewhat more likely to have a higher graduation rate.
Breaking down these same metrics by funding type, we see two distinct clusters for public and private-not-for-profit schools, as well as more variation among the few private, for-profit institutions offering majors in architecture and related fields. For public institutions, at higher net prices the graduation rate tends to be considerably higher. As their price points and graduation rates are both typically higher, this trend is not as visible to the eye for private not-for-profit institutions, although numerically there is a correlation.