Caution is necessary when interpreting the national data shown by default in the pipeline figure! As we describe in more detail here, there is no nationally representative dataset that allows for longitudinal tracking of candidates in the teacher pipeline—instead, each node in the pipeline is informed by statistics reported across both national (e.g., NCES, Census, Title 2, IPEDS) and state sources (Washington state West B statistics, Goldhaber et al, 2014), and, in some cases, our information is pretty spotty. Indeed, there are some nodes in our teacher pipeline tool that are not informed by data at all and are instead imputed from adjacent junctures on the pipeline. While this approach is likely to be less accurate than other studies that use state longitudinal data that provide a better state-specific view (Rucinski & Goodman, 2019; Wan et al., 2021), assumptions are necessary to provide a national picture of how diversity changes throughout the pipeline.
Another important caution is that the default assumptions for national rates may not apply well for individual states because they do not account for cross-state movement of candidates. For instance, teacher candidates may complete their college degree and TEP training in California and enter the workforce in Massachusetts—this would lead to national numbers overstating applications in California and understating them in Massachusetts.
Below, we describe the sources of information used in the pipeline figure: