data – Mechanical Sympathy

Bending People to the Data

Originally published on Dusty World in July, 2014:

The idea of data driven learning has become very popular. This isn’t surprising since data is beginning to drive everything. It becomes problematic when data is manipulated for ulterior, usually political motives rather than being understood in its own context.

It’s a complex series of events that have led us to this point. We’re living in an age of data where we are recording much of it for the first time. We mistakenly describe this as ‘creating information’, but we’ve always done that. What digitization does is allow us to save that data on a massive scale and then make connections in it we couldn’t before.

We’re not creating any more information than we used to, but we are recording it now at an unprecedented rate.

We’ve been experiencing this information forever. If I went for a ride on my motorbike in the 1950s I would have experienced roughly the same ‘data’ that I’d experience going for a ride now. The difference now is that the go-pro on my helmet and youtube means that data is saved and shared. We’re not creating any more information than we did before, but we’re recording it and allowing others to experience data now on an unprecedented scale.

This mass recording and access to data is a relatively new phenomenon so we should take care to contextualize it, but we don’t. We recognize that data driven methods yield results, but in our rush to enter this brave new world of data we happily ignore what doesn’t suit our goals and take other data out of context if it serves our cause. When politics or self-promotion drive data selection the benefits of data driven management are in doubt. Politics and self-promotion always influence data collection and presentation.

Since it is so much easier to record and share data we’re tempted to structure our activities around data creation rather than being present in the genuine experience. I suspect we’ll get better at this as technology becomes less invasive and allows us to capture moments with minimal interference. The evolution from TV to analogue video to digital video is a good example of this progress. But in the early stages of this evolution we’re still awkwardly focusing on data collection rather than genuine experience. Selfies at the Tour de France this year are an example. If you watch any live event where people are focused on recording rather than experiencing the moment you’ll know this is endemic. From the World Cup to the Olympics, the focus on data collection gets in the way of being there. This creates some interesting changes in experiential value. You now need to share the experience live rather than relating it after the fact. Being there is less important than your recording of being there. Every experience is one step removed.

Education is no different. Rather than focusing on ways to capture genuine experience in as non-invasive a way as possible, we create artificial situations that produce data for its own sake; standardized testing is a fine example. Rather than integrating literacy assessment into genuine experience driven learning, we create an artificial testing environment that is designed to produce data. Students and the complexities of literacy are minor components in that process. We then base management decisions on the corrupted data that is produced from these artificial situations.

If data collection is the point of the exercise then the data you’re producing is a reflection of the data collection process more than it is a meaningful analysis of whatever it is you think you’re assessing.

Technology needs to be pushed to produce non-invasive ways of collecting genuine data. Not only will this allow people to bask in the moment rather than ignoring first hand experience to create second rate data, but it will also serve student learning by focusing on the learning itself instead of the data gathering processes.

We need to stop bending the people to the data and start demanding that the data find us where we are, in genuine experience. In the meantime it is vital that we don’t blindly believe that there are absolute truths in data that is produced for its own sake with ulterior motives.

Data Exhaust

At a recent educational technology conference in Phoenix Constance Steinkuehler mentioned the term ‘data exhaust’ in passing to describe the numbers pouring out of testing. The idea of data as pollution has been with me for a while. The statistics I’ve seen derived from data in education have often been farcical attempts at justifying questionable programming. It’s gotten to the point that when someone starts throwing charts and graphs up in a presentation I assume they are hiding something.

.
Constance’s term ‘data exhaust’ had me tumbling through metaphorical implications. If the data we generate out of education is the exhaust, what are we doing when we turn the education system toward producing data exhaust for its own sake? No student will ever face a standardized test in the working world, it’s a completely unrealistic and limited way in which to measure learning let alone prepare students for the rest of their lives. Using standardized testing to measure learning has us revving the education vehicle at high rpm in neutral; we’re making a lot of smoke and not going anywhere.

Is data always useless? Not at all, but the tendency to find patterns and turn data in statistics takes something already abstract and abstracts it even further. That people then take these inferences and limited slices of information as gospel points to the crux of the crisis in American education. We end up with what we think are facts when they are really fictions that use math of lend an air of credibility.

Even with statistics and data metrics off the table, the idea of looking at the data exhaust pouring out of education as a way of directing future action demonstrates staggering shortsightedness. Education is not a data driven, linear or binary enterprise, it is a complex human one. We are not producing expert test takers, we should be producing well rounded human beings that can thrive in a complex, competitive, data rich century. No standardized test can measure that.

If you took your poorly running car into a mechanic and they just kept revving the engine harder and harder while watching smoke billow out of the back you’d think something was wrong with them, yet that is how American education is tuning itself. They then wonder why they aren’t scoring well in world rankings. If we want the education vehicle to take us somewhere we need to crack open the hood and take a look at the engine, but what is that engine? What actually makes the engine of education run well? It isn’t fixating on the data exhaust.

Canada has performed very well in world education rankings. We find ourselves able to keep up with some of the world’s best education systems, like Finland, and we do it at a much lower cost per student than the US has managed to. It looks like all that testing and data exhaust fixation costs a lot more than your students’ well being, it’s also hugely inefficient.

A well running education system focuses on pedagogy. It is what fuels it, it is what makes the system serve its students using the best possible learning practices. Pedagogy is a tricky concept, and it doesn’t offer simplistic solutions that digital technology companies can app-up, but it does give everyone, no matter how much they may disagree on the details, a common goal.

There was a lot of talk about coming together and pulling in the same direction over the Common Curriculum at this conference. We aren’t all on the same page in Canada when it comes to processes or how the system should run, but pedagogy is on everyone’s mind. Best practices have to drive education. Having standards isn’t a bad thing, but when you’re so fixated on the data exhaust you’re producing that you forget fundamental pedagogical practice, you’ve lost sight of what education should be in the data smog you’ve created.

In Canada we pay less and produce more by focusing on pedagogy rather than empty data gathering (aka: standardized testing).

via USC Rossier’s online Doctor of Education

Who Owns Your Data?

Tweets from the #edinnovation Summit

There was much hand-wringing over privacy and data ownership at the recent ASU/GSV conference. Serious people in designer suits explained that security is expensive, must be shrouded in secrecy and is never full-proof. Sounds like a great fear-sell for cash infused education systems (sic). Fortunately you can’t really oversee it either because that would be a breach of security; it’s a brilliant piece of hard-salesmanship with minimal oversight.

Listening to the urgency and paranoia in these discussions had me thinking about how privacy and data ownership are framed in digital environments. I’m the first to suggest we be cautious with how student data is shared, but the idea that anything digital can somehow be locked down and owned is ridiculous. I’ve heard ed-tech ‘experts’ go on about walled gardens at length. You can leap over any of those walled gardens with a simple screen capture or text grab, and then it can go viral on the wild internet. Trying to keep data in check is like trying to hold back Niagara Falls with a door, digital data is incredibly fluid, that’s why we can do so much with it.

The second tweet at the top sheds some light on how we are misframing data as an ownable entity. Like your appearance or your reputation, the data you give off is information you emit socially; you can’t own your data any more than you can own your appearance or your reputation. In all three cases you can, of course, influence the social outputs you’re giving, but you can’t own them. Realizing this would stop us from trying to control the uncontrollable and instead let us begin teaching students (and everyone else for that matter) best practices for creating desirable data emissions (there has to be a better word for that, digital reputation? digirep? virtual scent? vBO?).

If we begin to see digital data emissions as a natural byproduct of online interaction (and make no mistake, they are), then the idea that if we throw enough money at it we can make everyone safe gets questioned. You’re never going to hear education technology companies that market based on security suggesting this mindset, there’s no money in it.

The other side of this is privacy. The idea that we can suddenly have privacy now when we’ve never had it before is baffling to me. Before we became industrialized we were mostly geographically locked to where we were born. Without the ease of fossil fuel powered vehicles most of us never travelled more than a dozen miles from where we were born. Do you think anyone had privacy in those circumstances? You were a well established element in a fairly static society. The idea that you could shield your actions from the public eye only really began in the twentieth century. Industrialized anonymity was as much curse as it was something desirable. It was only in the conformity forced on us by industrialization that the illusion of privacy existed.

If the NSA and CIA can’t stop it, do you think that edtech companies can?

We are emerging from that drab, industrial anonymity caused by cookie cutter conformity. The digital world doesn’t expect us to be passive eyeballs counted as ratings in front of a TV screen, it demands interaction and individuality and it restores our voices. As we emerge into an individually empowered, digitally driven world of shared information, expecting privacy is a wish for a poisonous illusion that never really existed.

Privacy isn’t dead, it never existed. Prior to industrialization we had no social privacy at all and during industrialization we were dehumanized into bricks in the wall and given the illusion of privacy because we barely mattered as individuals.

When you’re online nothing you do is private, nothing you do is owned by you. Just as you can influence your appearance by good hygiene or your reputation by performing right action, you can affect your data-appearance by presenting yourself well, but ultimately any data you give off is out there and can spread easily, just like gossip.

The nascent study of digital citizenship addresses a lot of this, but like other digital skills it is an afterthought, never integrated into core curriculum. No wonder men in expensive suits can make lots of money preaching fear and convincing educators to spend their budgets on myth and innuendo. Perhaps one day educators will take the job of understanding and teaching digital skills seriously and ignore the snake oil.