In the field of education, isolating rules and processes to advance teaching and learning is widely perceived as central to academic and even social outcomes. In previous eras, policy and procedure adoption was largely influenced by social norms and the desires of the majority population because our schools are a microcosm of social conditions. For example, prior to 1954, the segregation of America’s schools was based on racial or ethnic group membership classifications, as was custom in the greater microcosm of society. In recent years, the shift in academic accountability has propelled data to the epicenter of educational decision making. In lieu of social norms, such decisions are principally induced by quantitative and qualitative data.

Educational Decision Making

According to Mertler (2014), educational decision making is a broad category to emphasize the processes educators follow to examine student assessment data and use the data to guide instructional and instructional-support practices. Mertler further asserts that educational decision making has been influenced by those he considers “old tools” like race and “new tools”. New tools include the data-drive science of teaching. While much of Mertler’s commentary focuses on the use of standardized tests as a primary mode of assessment, the utilization of “newer tools,” like portfolios, knowledge production, participatory artifact-producing learning, and student self-reflections warrant a wider examination and are certainly more representative of the way modern students interact with the world around them.

Data Mining

Despite the mode of assessment proposed by Mertler, his emphasis on data is obviously an important one. To further reduce the complexity of educational data decision making, I decided to take a look at data mining and discovered various definitions for the term. The meaning that provided me with the most clarity was offered by Urbina Najera and de la Calleja (2018) who indicated that “data mining can be defined as the process of extracting knowledge hidden from huge volumes of raw data” (p. 111).

Decision Tree

Within data mining exists the use of the decision tree, which is a widely used algorithm to show the sequence and structure of a decision problem. For example, tutoring is a process that has been linked to academic achievement, but on many college campuses, the selection of tutors poses a problem that, in turn, influences the validity of the student-tutor exchange. Thus, Urbina Najera and de la Calleja used decision trees to identify skills of college tutors to better match them with students.

For a better comprehension of a decision tree, I found Syncopation Software’s YouTube explanation most beneficial:

https://www.youtube.com/watch?v=MMHQ2gUMoNA

Educational Data Mining and the Classroom Teacher

Now that I have provided a general overview of the relationship among educational decision making, data mining, and decision trees, I feel responsible to highlight a process-like evaluation of a disparity of computational machine learning and related decision making. Data mining, overall, is a specialized science. Though certainly beneficial in its application, it is so focused in nature that the average classroom teacher, who is the most significant decision maker in the educational process, would have limited familiarity with creating a decision tree to isolate the “answer” to a decision problem. Far too often, America’s classroom teachers are not involved in assessment application selection or decision making nor are they regularly given autonomy to adapt decisions. However, the teachers are those who are charged with the instruction. With that said, how can we narrow down computational machine learning, data mining, and even the decision tree to help classroom teachers isolate student-specific—or at the general level cohort or group-specific--instructional questions?

If a teacher asks the question, “Is this read aloud culturally relevant enough to engage a group of sixth grade students?” the “optimal decision alternative” to the question would differ between a teacher who asks such in Chicago, Illinois and another teacher posing the same question in Miami, Florida. Moreover, with a consistent focus on student differentiation, the assessment (e.g., test, artifact, interview) that provides mastery evidence for one student would pose a challenge for another. Therefore, educational decision-making and data mining have to be drilled down and made meaningful, usable, and adaptable for the American classroom teacher.

References

Mertler, C. A. (2014). The data-driven classroom: How do I use student data to improve my instruction? Alexandria, VA: ASCD Arias.

Urbina Najera, A. B. & de la Calleja, J. (2018). Selection of academic tutors in higher education using decision trees. REOP, 29 (1), 108-124.