Measuring What Matters Part II

A new report will be released in California regarding the state of education. The report, prepared by 30 "experts" in the field, cites difficulties in sharing data, too many regulations and requirements for leaders, and ineffective methods for identifying effective teachers.

In a previous post I argued that we should be working to "measure what matters". In other words, if we want to know whether a teacher is effective we need to develop an example of "effective" and tools for determining whether the teacher is achieving that example. Teachers and leaders need to know where teachers on a scale (or rubric) and they need to know what to do next to improve. Years of service apparently can differentiate a 1st year teacher from a 5th year teacher, but after that there is little difference between a 5th year teacher and a 15th year teacher. Instead of seeing our longest serving teachers as the most skilled, we should create measures to see our most skilled teachers as our most skilled teachers.

Many Eyes is Addictive Data Visualizing Tool (it's social too)

Many Eyes, a product of the IBM Visual Communication Lab, is a wicked cool tool for seeing all kinds of data in a new way. Many Eyes says their goal is "to "democratize" visualization and to enable a new social kind of data analysis". In practice this means that all data uploaded to this site are public and all visualizations are public and can be commented on (like a blog).

Many Eyes is user-friendly and addictive. I kept loading new forms of data to try different views. I loaded school district demographics over time, speeches made by Margaret Spellings, and rap lyrics from Ice Cube and TuPac.

Imagine the potential in a K-12 setting where students could be challenged to collect data (primary collection or secondary) and then these data would be shared across the world. their classmates could comment, but so too could an expert in the field they are studying. This tool not only creates opportunities to see the world in a new way (literally), but also to collaborate and understand the world more deeply (or see it in a new way metaphorically).

Check out these examples I created:

Example #1: District demographics over time. Click the image here and drill-down (using the plus signs) to the demographics by school, ethnicity, and gender.


Example #2: A tree map of two years of demographic data reveals change in color. Check out the hover over Ajax features.

Example #3: Tag Cloud of recent testimony from Secretary of Education Margaret Spellings. Any idea what Spellings intends to focus on? Could you imagine students using this to compare the language in two poems, rap lyrics, books on the same subject in two different decades, speeches...or anything else? Imagine how engaged students would be to see the text they are analyzing come alive.

Great Design Works for Data Display Too

Excellent post over at Bokardo called Five Principles to Design By. In this post Joshua Porter identifies Five Principles that social web designers should keep in mind and these are easily extended to any school district or architecture designer that is thinking about data display.

First, technology serves humans. If the technology fails or the user cannot figure out how to get value from it, then the design is the problem. A simple concept that seems to be overlooked when considering how to display data to teachers. Second, design is not art. Art is to be enjoyed and design facilitates use. Third, designers do not create experiences, they create artifacts to experience. This seems to be akin to Kathy Sierra’s argument that serving our customers means that in the end it is about them kicking ass. Fourth, great design is invisible because it solves a problem and works well. We take it for granted. Fifth, simplicity is the ultimate sophistication. Distill the design to the absolute needs to solve the problem at hand, that is all.

This all applies directly to design of data systems for school districts to use. If the ultimate goal is for teachers to use the data system to analyze data and track student progress, then the design of the system must be teacher-friendly, focused on their experience, and simple enough that the user does not have to read the freakin’ manual.