Conducting national studies of schooling is notoriously difficult. There are nearly 14,000 school districts, over 100,000 schools, and the diversity of state structures, district administrations, and school types makes national research a real challenge. We have many unanswered questions about “what’s happening in U.S. schools” because studying the population as a whole is challenging.
Up until this point, the best nationally-representative research has been survey-based research, much of it produced by the federal government (which has a special capacity to compel people to do things. But technology offers some exciting new options, which I’m interested in exploring. For instance, in my own research, I drew a random sample of wikis used in K-12 settings from a population of nearly 200,000 wikis. I ended up with a sample of 255 wikis used in public schools, and they happen to come from 43 states. It’s not truly a representative sample of schools, but it’s a sampling methodology that is a promising way of examining national trends without sampling directly from schools.
A similar research approach is being pursued by the folks at Education Superhighway in order to learn more about school Internet speeds, and to be successful, they need your can help.
In general terms, we know that many schools don’t have the bandwidth to handle modern Web traffic. But how fast is the Internet in America’s schools? How does speed vary among different schools serving different population? We have some survey data from the federal government, but how school CIO’s answer surveys may not accurately reflect actual broadband speeds.
To get more accurate answers, Education Superhighway is running a National Speed Test at SchoolSpeedTest.org. It’s pretty straightforward: get on a computer from your school network, fill out a brief survey, and let the website run a brief speed test. If enough people in enough schools do the test, we’ll have a massive dataset on actual upload/download speeds from actual users in actual schools. These kinds of crowdsourced examinations of what’s actually happening on the ground in schools have tremendous potential to provide new views into the 100,000-headed hydra that is our national education system.
Go take the speed test, and let me know what it tells you here in the comments!
For regular updates, follow me on Twitter at @bjfr and for my papers, presentations and so forth, visit EdTechResearcher.
**A brief methodological note: one problem with surveys of this kind is that it is a “convenience sample;" that is, the survey is being taken by people who happen to encounter the survey rather than by people systematically selected to take the survey. It may be, therefore, that the schools where the speed test is taken are somehow different from the schools where it isn’t taken. For instance, if most of the tests are conducted in schools in wealthy suburban schools, and few speed tests are conducted in poor rural and urban schools, then the “average Internet speed” found among sampled schools may be a biased estimate of the actual national average. Fortunately, we have extensive demographic information about every school in the country, so we can compare the Education Superhighway survey sample to the actual school population, and we can determine how the sample differs from the national population of schools, and if necessary adjust our interpretations of the findings. This may prove to be an example where, even with a convenience sample, we can draw fairly authoritative conclusions about the population because we have extensive data about the entire population and—through crowd-sourcing—we have the capacity to draw a very large sample.