During the 2003–04 school year, sweeping revisions to Vermont's student census included the modification of the state's race and ethnicity data collection standards. To more closely align with the 1997 guidelines, the state adjusted its standards to allow multiple race selection in the collection of student data.
To permit the selection of more than one race, the state merely expanded an existing feature in its old collection system. Previously, schools reported, for each student, either “Yes” or “No” for Hispanic in addition to a single race. To allow the reporting of multiple races following the new standards, the state simply added a “Yes” or “No” option to all of the race categories and now allows respondents to select “Yes” for multiple races. The data for each student must include either a “Yes” or “No” answer for Hispanic and all five race categories. The data collection application does not allow submission otherwise. Furthermore, to assure that ethnicity and race data remain distinct, the system requires that if Hispanic is marked as “Yes,” at least one race category must be checked “Yes,” as well.
In Vermont's education data warehouse (EDW), there are seven race and ethnicity fields and three different “race unknown” categories, coded as follows:
|1||American Indian/ Alaska Native|
|4||Native Hawaiian/Pacific Islander|
Since race and ethnicity data must not be missing, observer identification is performed for all students who do not self-identify. The BLANK and NULL fields accommodate exceptional cases, such as privately-funded students at independent schools. ANY_OTHER is part of the system's quality assurance process and is activated when an invalid value is entered in the field – it is an error message rather than a valid entry.
In the EDW, each race marked with a “Yes” is counted as 1 count. At this point, you may be asking yourself, “Hey! Since some kids receive more than one race count, doesn't that mean that the total race count would exceed the actual number of kids?” Well, thanks to the magic of the state's EDW, this problem is averted. In practice, a Hispanic/White student is coded as 5 | 6 from the table above. In the EDW, codes 5 and 6 are both maintained, but are separated by “pipes” in the system and stored in different “buckets” so that the codes are discrete and can each be independently accessed for querying. That way, if a student's record includes the codes for both Hispanic and White, the EDW “sees” the child as only one student, as it “knows” not to count more than one count per student for the total student count. On the other hand, if a query is made for the total number of students that are White or for those who are Hispanic, that particular Hispanic/White student would, however, come back as one count in each of the race or ethnicity totals returned.
Currently, while students who select more than one race are counted as “multiracial” in the Vermont system for assessment regardless of their ethnicity, for federal reporting, Hispanic multiracial individuals are counted as “Hispanic.” In the future, Vermont will create a new Federal Racial Grouping attribute that aligns with the new EDEN requirements when they are implemented.