GOLDEN MEANS: DISCOVERING THE ACADEMIC POTENTIAL OF OUR CHILDREN
PART I
by Miriam J. Masullo, guest columnist
Part 1: The problem with the assessment of academic potential
Just as not all that shines is gold, some things that do not always shine are as precious as gold. So it is with the minds of our children. There is a lot of value buried in the depths of data warehouses being built by corporations, but it is in the depths of time yet to come that the real gold lies, hidden in the future of our youth.
Certain problem solving styles and cognitive behavior reflect the academic potential that leads to success in college. Often, students who are bypassed by traditional recruiting activities have been known to demonstrate such potential. In such cases, potential that is unrealized, wasted, lost.
Alternative assessment methodologies are just beginning to touch upon the unknowns in student achievement and potential. Many aspects of those programs rely heavily on observation and human analysis of the data generated via observation, making them labor intensive. These kinds of complex analyses, however, often yield more accurate results for under-served populations than do traditional assessment practices. Some mentoring programs that emphasize gathering of data for comparison in various groups, selected by specific characteristics, tend to provide clues for identifying and nurturing promising candidates who are usually bypassed because of early poor academic performance. Such approaches that account for unequal early education preparation, merit consideration.
Imagine an educational system where all students can realize their full potential. A system in which gender and race play no roles in determining performance and expectations. Imagine equal educational opportunity driven by talent and achievement. Those who evaluate students make their future. How we evaluate them decides the future of our nation. We have a national responsibility to ensure that every student is correctly assessed for all the potential that is our nation's to harness and nurture. Assessment practices, more than any other education responsibility, require our national attention. Industrial growth and productivity gains in a global economy are expected to increasingly depend on improvements in our ability to apply and analyze new knowledge. Emerging data mining technologies are meant to do just that. But they should go further. Help address the additional challenge of not indulging in the luxury of bypassing any human potential.
There is an increasing and compelling national need for making the identification of academic promise a priority. We need to enhance the quality and quantity of the American work force by drawing qualified candidates for engineering and high-tech careers from our nation's entire population of students. Properly harnessed, advanced data mining techniques can have considerable impact on alternative assessment practices, and indeed on the discovery of intellectual identity for all students. But, to begin to apply these concepts on a broad scale means that we must use the same strategies for change that the nation's corporations are using to remain competitive. Standardized data analysis of test results is no longer sufficient. In a technology and information driven society, a preferred approach is to combine advanced technology with methods for alternative assessment to identify human potential.
In Performance Based Assessment and Educational Equity, Darling-Hammond points out that "efforts to raise standards of learning and performance must rest in part on strategies to transform assessment practice. Normative and multiple choice tests fail to measure complex cognitive and performance capabilities." A belief shared widely, because, the ability to use and construct knowledge, synthesize, and apply information is becoming increasingly valued in education. Assessment and evaluation practices must, absolutely, explore the complexity of problem solving skills, styles and competencies that predict academic success, not just measure it at existing levels. The process of knowledge discovery of student potential would require the detection of patterns, which permit these kinds of alternative assessment practices. The current global approach to assessment relies on elicited data, by means of testing, rather than on discovery of potential. The result is a number that represents a classification of students by performance in a test. This approach is exclusive. We need a more inclusive model.
It may be that the place to find this, more inclusive, model is in Howard Gardner's much-acclaimed theory of multiple intelligences which challenges the very premise that there is a single human intelligence and that it can be measured by means of a test and ranked by a single number. According to Dr. Gardner, eight integrated intelligences compose the measure of human potential. Taking this theory one step further Barkman, in The Building Tool Room, has suggested that the ability to recognize and classify intelligence comes from a natural expertise in identifying patterns.
Data mining is about precisely that: The automatic discovery of trends and patterns in large amounts of data resulting in knowledge discovery. Hence, Barkman's thesis offers a clue as to what data mining technologies might be able to do for student assessment.
However, what can be learned about students by the use of data mining techniques is subject to the data being collected. Current assessment practices do not elicit data that can be useful in the discovery of new knowledge about student potential but are more focused on assessing performance in a test. Data mining techniques will enable one to get answers to specific and general queries (about student performance) but that can only be done after appropriate data has been collected and is present in the assessment process.
It follows then that assessment practices must be changed to extract the
kind of data necessary for identification of talent by examining those
problem solving skills and cognitive behaviors that many educators believe
are significantly correlated to performance. Recording this class of data
requires paying attention to and documenting examples of verbalization of
concepts, and of the framing of questions. It requires noticing how a
student engages other students and faculty, uses available resources, tests
ideas, produces ideas, structures solutions and ultimately solves problems.
It includes detecting and documenting the student's ability to construct
new knowledge and of other cognitive behaviors known to contribute to high
performance and that hide with them the promise of academic success. If
those data are collected, the discovery of new knowledge about performance
using the emerging data mining technologies can identify those highly
qualified individuals who are missed by standard assessment practices.
---
The concluding portion of this article will appear in the next edition
of D S * .
---
For more information, see http://www.virtualgold.com