Multiple growth models for TeacherCandidateStudentGrowthMeasure

We often have multiple growth models associated with a single teacher candidate, for the same set of students' scores. One of those models might be SGP (student growth percentile), one might be TLSD (teacher-level standardized units), etc.

For this same domain of TeacherCandidateStudentGrowthMeasure, we would benefit from a "model type" field. This would be especially helpful for the purpose of querying, through which we would never want to compare the results of SGP vs. TLSD, as the units/range are completely distinct, rendering any such comparison incomprehensible.

SAA - Originally submitted as a comment on TPDMX-57