I delivered my most “brilliant” lecture early in my teaching career. My junior English class was reading Nature by Ralph Waldo Emerson, and both the concepts and the language proved challenging. I, myself, was not a huge fan of this extended essay when I was in high school, but as I re-read it as an adult, I was fascinated by Emerson’s use of imagery to shed light on the complex spirituality of Transcendentalism.
I spent entire class periods passionately explaining Emerson’s ideas that matter was more a phenomenon than a substance and that god was not a separate entity above us, but a presence that permeated our being – I used student-friendly language and drew elaborate pictures and diagrams on the board. From where I was standing, the class watched in silent awe and admiration, dazzled by each pearl of wisdom I bestowed upon them. When it came time for the essay test, I watched in horror as students stared blankly at the questions before them – questions for which I had given explicit answers in my enthralling lectures over the previous week.
Why Use Data to Drive Instruction?
This early lesson in the gap between what we “teach” students and what they “learn” should be familiar to any classroom teacher. Good assessment shows us what students understand and retain; while performance-based and other subjective assessments are playing an increasing role in how we understand and support student learning, the importance of data can not be discounted. In an interview with Edutopia, Grant Wiggins (co-author of Understanding by Design) explained the importance of objective data to help “triangulate” (expand, verify) the information about student learning that is gained through performance-based assessment.
A 2007 study by NewSchools Venture Fund supported the idea that when educators use data to drive instruction, they measurably increased student achievement. In order to effectively differentiate instruction, it is crucial to move beyond the idea that giving a student an 87% is a meaningful assessment and that writing comments on a paper is meeting individual student needs. Two students may both score 87% on a quiz while making different types of mistakes on different types of questions. Writing ‘r/o’ in red on a paper does not mean a student understands what a run-on is or how to fix it (even if you assume they’ve read the comments).
True data-driven instruction requires that we assess not only what the student has done right or wrong (Step 1), but the skill or knowledge gap that has caused their success or failure (Step 2). Once we have figured out the obstacle to understanding, we respond with a plan that targets the gap or gaps (Step 3).
Step 1: Gather Data
Whenever we start a new project or implement a new habit, the first step is always the most intimidating. When educators try a new technology, this step can be exponentially more intimidating. As teachers, we are used to being experts, and many of us are reluctant to try something out on our students until we completely understand how it works, and we can anticipate problems that may arise. Just remember, you are already generating data – any time students answer a question on a worksheet, quiz, or in a discussion, it tells us something about what they know. With data-driven instruction, we are simply using technology to preserve and organize the data in a way that we can understand and interpret.
Therefore, the most important part of Step 1 is ensuring students complete the activities. I was amazed at how much I learned about my students’ homework habits once I could view the amount of time they were spending on assignments and when they were spending that time. On several occasions, I have learned that a struggling student was spending a lot more time than was apparent (or a lot less time than they said they had!)
When incorporating data-driven instruction into a class, it is helpful to know that if we take a leap of faith, the benefits will be worth the effort. It is difficult to understand how data will help you differentiate learning until you see actual results from your actual students. Every tool I’ve used (Khan Academy, NoRedInk, Vocabulary.com, GoogleForms, etc) offers a way to view sample data sets, but the data just does not have the same resonance if it is not about your students and your content.
Here are some ways to help you and your students take the first step toward implementing data-driven instruction:
- Create time in the schedule – any time you add a new component to your class, you will need to eliminate something, or you and your students will become overwhelmed. Try to find something that duplicates what the online platform can do – I got rid of our textbook and have never looked back!
- Complete the work in class at first – You will need to remove obstacles that can prevent students from completing the activities. By having them work on the platforms during class time, you can make sure they have access to devices and the internet, and they have a clean well-lit space to work in. They will also have you there for tech support, as some students will need help logging in to and navigating websites. Once they are comfortable, most students will be able to complete activities outside of class as well.
- Grade students on time spent (or questions answered), not on correct answers – A level playing field is core to the idea of blended learning, so incentivize students to work at their own pace and on their own level. By grading for time spent, you will gain buy-in from students who have low confidence in the subject area. It also fixes the misguided incentives of typical grading systems, in which students are punished for incorrect answers, rather than being encouraged to persist through challenges.
- Make it fun! – Many blended platforms create leaderboards (Khan sends a weekly e-mail and Vocabulary.com posts a leaderboard on your teacher dashboard). Each Monday, I play The Theme from Hawaii 5-0 and announce the week’s top achievers (in terms of minutes spent), game-show style. In class, students work in leveled groups and earn points when their group members can all complete an activity.
It’s essential that when you implement a data-driven program, all students feel that they can achieve success – the whole purpose is to differentiate learning – that means both differentiating the content and finding a way for all students to connect with the new approach. Students who are above or behind grade level tend to disengage from classes that focus on punitive measures to incentivize completion of work (especially when the work is not differentiated to their level). Adding competition must be done carefully: points should be awarded for completion of work that is appropriate for each student (in other words, the advanced kids have to work just as hard for each point as everyone else).
By adding elements of gamification, I have seen significant increases in homework completion and in student enthusiasm. (Be aware that simply awarding points does not mean you’ve gamified your class). I’ll cover more on gamification in a future post, but here’s a great blog post from TeachThought to get you started).
What do you think?
Share your experiences with data-driven instruction, or ask a question below. We’d love to hear from you!