Student demographic data is too broad to show how individual subgroups are doing in school
The Washington State Office of the Superintendent of Public Instruction publishes a data manual each school year that lays out the information collected on students at the almost 300 school sites across the state.
The document is 127 pages long, and its appendices add another 61 pages to the total.
It lays out coding information for 244 countries, 327 languages, 10 ethnicities and 56 races, but that’s not the information that decision makers see. Instead, 32 Native American tribes become “American Indian/Alaskan Native,” nine countries become “Native Hawaiian/Other Pacific Islander” and 16 groups become “Asian.”
That’s a problem for people like Sharonne Navas, co-founder and executive director of the Equity in Education Coalition.
“If you don’t disaggregate the data, it ends up being the average of all things put together,” she said. “You have kids doing really well put in the same pot with kids doing not so great. It becomes really unreliable data.”
Averages are useful, but they are terrible at dealing with outliers, the individual numbers that break from the mold and end up considerably higher or lower than the majority.
Most of the time, it’s all right to discount those outliers as exceptions that make the rule. But you need to be comparing things with the same characteristics, and that’s where it falls apart with demographics.
Lumping together nationalities and races is a big deal, such as for people listed as “Asian.” The continent is the world’s largest, including everything from China to Palestine. These nationalities have a range of histories and experiences in the United States and King County, which has become increasingly diverse since 2010, according to the King County Office of Performance, Strategy and Budget.
Averaging information such as school test scores hides when some groups are doing well and others are struggling. It becomes difficult to target help toward the groups that need it, said Brianne Ramos of the Commission on Asian Pacific American Affairs.
“With funding it’s important to know what communities are there and then we know who to serve,” she said. “Asian, Asian Pacific Islander you run the risk of hiding those disparities.”
Before the 2010-2011 school year, the State Board of Education data showed that 65 to 70 percent of Asian students were meeting academic English and language arts standards. Starting in 2010-2011, the federal government began recognizing “Asian” as distinct from “Pacific Islander.” Suddenly, “Asian” students were meeting standards closer to 70 percent of the time, whereas “Pacific Islander” students were meeting standards a little more than 45 percent of the time. When students were allowed to self-identify their race or ethnicity, the range of academic achievement became even broader.
Each group of students still tended to do better than Pacific Islander students, but that ranged from more than 85 percent of Taiwanese students meeting standards to a little more than 55 percent of Laotian students doing so.
The integrity of that data is suspect, however, because while it now gives many options for races or countries of origin, it isn’t built to code for biracial children. If a student is both Vietnamese and Korean, for example, checking both of those boxes shows up in state data as two individuals.
“There’s a lot of doubling up,” Navas said.
It creates a situation where staff at school sites are the only ones who know with absolute certainty who they are educating and how well they’re accomplishing that goal.
Part of the problem is that data-collection standards are driven by the federal government, which uses broad categories in massive efforts such as the United States Census that takes place every 10 years. Even the feds are coming to understand the limitations of those categories, said Alisha Coleman-Jensen, a researcher with the United States Department of Agriculture Economic Research Services.
The organization plans on putting out research looking at food insecurity among Hispanic subgroups who experience the economic realities of the United States very differently.
Although there tends to be less variation amongst the Hispanic and Latino populations in Washington than in the Asian groups, more diversity is starting to show up in groups of African descent, even further highlighting the need for disaggregated data.
The lack of data may sound like a small problem in a state where education is so poorly funded that the Washington Legislature is getting fined daily for not putting enough money into the school system, but in fact that makes it even more critical that decision makers know where the need really is.
“At some point, I think we have to have that very hard conversation about how schools with large numbers of kids of color are not getting as much money as they need,” Navas said.