Data Driven Dilemma
Are you driven to distraction by student data? Is your school working on data walls, data-driven instruction, and using data coaches to help teachers learn and implement this? Are your teachers using data-driven differentiation, creating data-driven classrooms, and using student achievement data to support instruction? How much time, energy, resource, and faculty cognitive load have you spent on this, and is it working? Is it doing more harm than good?
In the ideal classroom, data-driven anything makes sense. The books on the topic have valid, meaningful suggestions and in ideal cases, fine results. Using short learning cycles, using data to assess students, diagnose difficulties, and prescribe targeted interventions to help every student is good practice.
However, the results don’t seem to be matching the effort. Even in Vermont, where the majority of elementary classes are smaller than 20 students, literacy and math scores do not show significant increases compared to the resources applied. This is not the fault of the teachers, the principals, nor their caring, effort and desire to help students. It is not the fault of the curriculum, nor the professional development efforts. It is simply that the challenge is too great for traditional educational models.
To meet the challenge in a self-contained classroom, an elementary teacher would need to:
be expert on multiple elements of standards
be expert on multiple years of standards
be expert on multiple methods of teaching (e.g. phonics v. whole language)……..in math and ELA…..
and have the time to focus on special interventions for how many students...in how many subjects?
The diversity of incoming students, their prior learning, and other learning variables (age, home support, etc.) mean that even if all of them were on the same page at the start of the school year, the variance would increase week by week. As one teacher put it, “I was using personalized instruction in math and by the end of the first month, the students were in twenty different places.”
There are other challenges -- frequent testing takes away from instructional time; continued mediocre results has a deleterious effect on teacher self-esteem and attitudes towards students; learning to analyze data and interpret results require data wonk skillsets that are not typically part of the teacher’s portfolio; and time spent on learning takes away from the meager professional development time most schools allocate.
While there are solutions in different school models, such as schools that have 10%-20% of faculty time weekly to spend on looking at student work to focus on improvement, these solutions do not seem to be taking hold across most of the public school sector.
There is another solution on the horizon, and it will present opportunities for the educational market. I see two possible directions:
Dystopian - Use an individualized AI learning space similar to what Spock experienced in the most recent version of Star Trek. While that is an extreme example, a personal AI educational tool may soon be available to parents directly through smart speakers such as alexa and google home….these tools are not likely to be marketed to schools due to student privacy issues, but parents who can avail themselves of these tools will want to be sure their kids have access as a competitive advantage.
Utopian - Use an AI computer-based adaptive testing/learning system to individualize basic assessment, prescription, and teaching of critical literacy skills (for example reading by the end of grade 3); provide results for the teacher to use to identify whole group and small group instruction; let the teacher focus on the human side of education (e.g., empathy), integrating into deeper learning (e.g., wicked problems), and focusing on subject areas such as science, civics, and social studies that often get left aside.
If we really want all students to be literate by grade four, then it’s time to allocate resources to systems that can help every student with literacy, rather than expecting to transform all teachers into data wonks. The costs of continuing to expect teachers to become data-driven educators are too high