Consider whether we can accommodate "abstract" assessment periods

From George @ Wireless Generation:

“Abstract ” assessment periods. AssessmentPeriodDescriptor works well for concrete periods associated with an individual assessment, but less well for abstract periods associated with an AssessmentFamily. For example, the AssessmentFamily DIBELS uses three periods: Beginning, Middle, and End of Year, or BOY, MOY, and EOY. It is not possible to define a BeginDate and EndDate across years, but without that the data model does not provide a way to represent the ordering of these periods within the school year. Similarly, the precise windows (BeginDate and EndDate values) may be different for related Assessments within a family, e.g. the DIBELS BOY window for Kindergarten might be a bit later than the DIBELS BOY window for first and second grade. Even so, it would be desirable to recognize the notion of an abstract BOY window that extends across all DIBELS grade levels and school years.

Relatedly, many assessments are administered on an annual basis, such as state summative tests in ELA and math. In this case, the year of the test is, in effect, the period. The specific window may vary, e.g., from grade to grade, but it is still meaningful and useful to talk about the “2012” New York State tests, across grades with different administration windows.

MSDF/DLP assessment:

We need to think about this further.