Beyond Standardized Tests: Student Survey Data in Matrix

We spend a lot of time in Matrix working on standardized test data. There is, after all, quite a lot of it, and almost all of our Matrix clients have standardized test data in our systems. Even so, I always enjoy getting a chance to work on data the districts are generating themselves. One of our clients periodically takes surveys of their students' experiences and attitudes, and they approached us about bringing this data into Matrix, so it would be available alongside the rest of their data.

There were different sets of questions for elementary, middle, and high school students at each survey administration, and though most of them were of the “Strongly Agree” to “Strongly Disagree” form, there were also questions asking about duration, frequency, or specific elements. We were able to teach Matrix how to distinguish the different answer scales and display the data as a grid or a chart, with all the filtering and drill down capabilities our Matrix clients are familiar with.

It's amazing how displaying data in a more engaging way can change the way you experience it. None of the information in these surveys was new to the client, but the first time they saw it in Matrix as a chart, it prompted a real discussion on policy between the administrators. That's a big goal of any data-driven interface: help users see where there are interesting questions to ask, then help them answer those questions and make decisions.

AP Predictor: Using Data to Drive Decisions

The AP Predictor report in Matrix is designed to identify students who are likely to do well in AP courses, so staff can be sure those students know about their AP course options. The predictive element is taken from research published by The College Board showing a strong correlation between PLAN scores and scores on specific AP tests. They provide a list of AP tests and their associated PLAN subscores and cut scores. For example, according to The College Board, students whose PLAN Math and Science scores average at least 26 have a 75% chance of scoring a 3 or better on the AP Microeconomics test.

Of course you could do this math by hand or in a spreadsheet, but your PLAN scores are already in Matrix, so why not let Matrix do the work for you? The AP Predictor report knows which PLAN subscores go with each AP test, and automatically calculates the relevant average when you select an AP test from the menu. You can filter by student name and minimum AP success category (e.g. only students with a 50% chance of scoring 3+ on the selected AP test), and sort by any of the scores.

Coming up with ways to display historical data that make it easier to absorb is a cornerstone of Matrix, but it's always exciting to build an interface that directly helps educators make decisions about the future.