Insights

My team’s story is about technological innovation. In 2004, when I began my profession at Cambium Assessment, all standardized tests were taken on paper and responses in bubble sheets were scanned into machines, then extracted for scoring and reporting. In the same year, our technology team had about 15 software engineers and technical project managers. We designed online practice tests to help students practice for their summative tests and built reports that sourced student test data from scanned systems.

However, around the same time, something radical was happening in how the world was accessing the Internet. For the first time ever, more and more schools in the U.S. had access to broadband internet, and shortly afterwards, they got wifi. The world became more silent as the distinct tones and beeps of a 36kbps dial-up modem were a thing of the past, but became busier and more connected.

Our Cambium Assessment team leaned into this new connected landscape and created a highly-scaled online testing platform that included everything from registration to test delivery to scoring and reporting. To date, we are serving over a third of students in grades 3-8 in the U.S. and have delivered 100.6 million online tests in the 2021-22 school year alone.

The relentless search for improving student learning continues to be at the core of everything we do. Here are three lessons I’ve learned:

1. Listen. Observe what your users, your teammates, and your competitors do, and then think about how to make your systems and processes better. Challenge your habitual perspective by observing what they do, mistakes they make – and equally important, what they do not say – by doing so, you just might identify a real-world problem. For example, in 2014, even though we delivered adaptive assessments to the general student population, we still delivered Braille paper tests to students with visual impairments. We identified a gap so we built an adaptive Braille test for students.
2. Timing matters. Don’t wait for the eureka moment to come up with the perfect idea. A real problem can strike at any time, and you should know how to address it with your product. Otherwise, it may either be solved by your competitor or become a bigger problem. Come up with a solution right then and there, prototype it quickly, then evaluate if it solves the problem. For example, in 2019, states required flagging of student responses that contain potential harm. Assessment companies were solving this by having hundreds of humans look at millions of responses and indicate which ones are alerts. In the meantime, artificial intelligence (AI) methods have become more reliable while more and more students were testing online. So, we built Hotline – an automated AI alerting system that has scanned 40 million responses alone in the 2021-22 school year. Hotline identifies student written text as an alert vs non-alert, and then automatically sends the alerts for humans to verify and escalate. This process takes minutes, not weeks or months, potentially helping the well-being of students and teachers at the time it’s needed most.
3. The innovation needs to work and it needs to work well. The assessment industry is a tricky one. On one hand, the idea of using science and technology to continuously understand students and measure what they know is an attractive prospect. Innovation, prototyping, and experimentation are welcome initiatives. On the other hand, if a feature is built incorrectly and student scores on high-stakes tests are impacted – well, that’s a bad thing. So the innovative idea needs to be built reliably, securely and in a scalable way because it has extremely tangible, real-world impacts on students.

Keeping these three lessons in mind – in unison – can help make your innovation a meaningful one. So just like what any writing test would look for, I hope these lessons are succinct, coherent, and cohesive, but most importantly, inspire you to innovate.