Note: Elliot Sanchez, the founder and CEO of mSchool, is guest posting this week.
If you’ve considered the benefits of personalized instruction and are prepared to measure student growth, the next step is to define the student needs that you want to address. As soon as we start to define those needs, we’re making decisions about what information is important and what can be left out. Decisions about data are fundamental to this process, but, from my recent conversations, school leaders often feel like they’re drowning in it.
When we talk about personalization of any sort, the assumption is that the system we’re using changes in response to our needs. Given the ubiquity of social media newsfeeds, it’s certainly a familiar concept. Introducing personalization into the classroom can be similar, but it necessitates protecting student information first and foremost. For many years, the protection of student data referred mostly to grade books and report cards, but adjustments are being made as students spend an increasing amount of time online in personalized software environments.
The bipartisan Protecting Student Privacy Act, introduced this month by Senators Edward J. Markey (D-Mass.) and Orrin Hatch (R-Utah), would amend FERPA (the Family Educational Rights and Privacy Act) to slow the propagation of student information without parental consent, curb the use of student data in commercial applications, and secure data held by private companies. The legislation has been praised for its balanced approach, but schools would still need to consider these changes when designing their student experience.
So how can schools ensure that students stay protected and still see the benefits of personalized instruction? Two key recommendations:
1) Know which data is being shared and how
2) Separate good data from bad, and get rid of the bad
Know which data is being shared and how
We’re accustomed to treating grades and academic information confidentially, but educators often forget just how much information about students is contained in school records. Most students’ names, addresses, race, sex, and even financial information can be found in student information systems. That information can be used to de-anonymize student data from much larger data sets, like learning games that students find with a search engine.
Online grade books, curriculum planning tools, student resources--all told, there may be dozens of places where student information could potentially be compromised. Being clear on which programs you’re using, what information they have, and how that information is used may seem basic, but given the realities of the classroom, it may not have been a priority in the past. Starting with a basic information inventory can give you better insight into unnecessary risks that might have been overlooked.
Separate good data from bad, and get rid of the bad
Can data really be “bad?” Maybe that’s a bit dramatic, but data is only good if it’s meaningful and helps solve the problems you’re focused on. One example: Nate Silver’s book The Signal and the Noise points out that, for 28 of the 30 years from 1967-1997, when a team from the AFC won the Superbowl, the stock market had a down year. When an NFC team won, though, the stock market rose. The odds of that correlation happening purely by chance are less than 1 in 4 million, but sure enough, it happened.
I mention this story because when we have enough data about students, there are an incredible number of ways to dissect and analyze it, but not all of them are helpful. As the amount of data grows larger and larger, even unlikely patterns are occasionally found.
In our programs at mSchool, we start with millions of unique data points for each classroom, but only a tiny fraction of those ultimately have value. By the time we’re presenting teachers with actionable reports, we’ve removed close to 99 percent of the data we started with. While some may advocate for holding on to the data because it may be useful in the future, summarizing and cleaning what’s left ensures that there’s less data to worry about being misused in the future.
Ultimately, data is only a tool, one that is as helpful or unhelpful as you make it. School leaders should understand the data that’s being generated with any new personalized learning initiative, and teachers will need to make additional time for analysis. Getting started can be daunting, but the insights you can glean make the effort worthwhile.
The opinions expressed in Rick Hess Straight Up are strictly those of the author(s) and do not reflect the opinions or endorsement of Editorial Projects in Education, or any of its publications.