Tuesday, March 05, 2013

Explaining inBloom in Terms of a Digital Scale

I've got a turn of the century digital scale that estimates body fat percentage as well. If I want to track that over time, I have to manually write it down or enter it somewhere. The new models will communicate that wirelessly to my PC or phone, and they are integrating more data sources, like a pedometer.

We can imagine it would be cool if we could electronically share that data with selected health care providers if there was a good reason to (maybe we can already).

What if you could get a scale that would automatically send your weight and any other data it gathers to a centralized system that tracked all health-related data collected during your life? Would that be cool? What if you couldn't turn that feature off?

What if you had to use that scale and upload every reading as a condition of getting government-funded health care?

What if the data went to a non-profit organization set up by Rupert Murdoch's News Corp and the Gates Foundation?

What if they provided your weight data to private companies that are trying to do research on your personal habits and create and market products based on that data? And they charged your state government for doing this (after getting them hooked on a free sample)?

As far as I can tell, that's an accurate description of how inBloom is supposed to work with your child's educational data. Everything measured, quantified or recorded about every child will go into the national data set.

If I'm getting something wrong, please let me know.


Frank Catalano said...

From what I understand, there's no "national data set." Every entity using inBloom uses a separate "copy," or instance, of inBloom's database software. The data set stops where the district sets it.

Kind of like an office installing Microsoft Word on individual laptops. Everyone has their own copy of Word with their own documents (same architecture), but the documents aren't co-mingled. In inBloom's case, the "laptop" is secure cloud-based storage.

Perhaps someone more technical could explain it better. But it isn't, as I understand, a single data bucket into which all student data from all sources is co-mingled. The buckets are kept separate.

Tom Hoffman said...

The one thing we know for sure is that they aren't trying very hard to make it clear at this point.

At this point, technologically, it is less clear in general what one "bucket" would be anyhow. Is Google's index one big bucket?

I suspect that inBloom would like to have everything in one bucket insofar as they'd like to be able to run queries, etc. on the entire data set at once, and if they can get away with it, that's what they'll do.

Frank Catalano said...

The queries are theoretically run by the district for its own reporting using applications the district approves that connect through the inBloom APIs, and are not run by inBloom itself.

Google is supposed to be a data set of the entire web. inBloom is supposed to be a data set of a specific (e.g., district) student population.

I agree that inBloom has (when it was Shared Learning Collaborative), done a pretty poor job of communicating how this all works to a non-technical audience. Which is most of us. That's what caused me to start digging into it in the first place, to understand it myself. We'll see if inBloom does better.

Tom Hoffman said...

It seems to me the intent is to have one big system that inBloom's non-school clients can do large-scale data mining on -- particularly if they are screening out the personally identifiable data.

I think the whole point is they want to be able to ask stuff like, out of all the students in the country who received intervention X, is there a difference in outcome by income status of the student? Stuff like that, that's why they want one bucket.