×

Steve Nelson: Testing Is Profitable, but Not for Students



Sunday, May 10, 2015
Measure the wrong things and you’ll get the wrong behaviors.

This simple statement succinctly characterizes why the American education system continues beating its head against the wall.

Education reformers and so-called policy experts are constantly collecting and analyzing data. Many of these experts are, not surprisingly, economists. It’s not for nothing that economics is sometimes called “the dismal science.” The hostile takeover of education by non-educators is filled with intelligent sounding phrases: “evidence-based,” “data-driven,” “metrics and accountability.” At every level of schooling, mountains of data are collected to inform “best practices” based on the alleged cause and effect implications of data-based instruction and the feedback gleaned from tests.

It is not coincidental that the education policy and reform business is highly profitable. Public education is estimated to be a $600 billion to $700 billion market. Those who drive the measuring and testing industry are first in line at the trough. Pearson Publishing, for example, has its tentacles in nearly every school district in America. All the iterations of reform — No Child Left Behind, Race to the Top and, more recently, the Common Core — are driven by (and driving) the collection and interpretation of data.

Throughout education, an increasingly rigid, closed loop of assessment is systematically making schools worse: define things children should know or be able to do at a certain age; design a curriculum to instruct them in what you’ve decided they should know; set benchmarks; develop tests to see if they have learned what you initially defined; rinse and repeat.

This narrow, mechanistic approach to education does not correspond to the reality of child development and brain science, but the metrics and assessment train charges down the track nevertheless. So what’s wrong with that, you might ask? Isn’t school about teaching kids stuff and then testing them to see what they’ve learned? In a word, no. It simply doesn’t work, especially with young children.

As Boston College Professor Peter Gray wrote in a recent Psychology Today article, “Perhaps more tragic than the lack of long-term academic advantage of early academic instruction is evidence that such instruction can produce long-term harm, especially in the realms of social and emotional development.”

“Direct instruction” does increase scores on the tests the instruction is aimed toward, even with very young children. In the context of these studies, it is the explicit teaching of skills, such as phonics instruction, worksheets with computation problems, lectures and other direct ways of teaching precisely those things that are expected to be subsequently recalled. This self-fulfilling prophecy is not surprising. But multiple studies also show that the gains in performance are fleeting — they completely wash out after one to three years when compared with children who had no such early direct instruction.

Wash out is too kind.

A comprehensive study of kindergartens in Germany revealed, as Gray writes, “Despite the initial academic gains of direct instruction, by grade four the children from the direct-instruction kindergartens performed significantly worse than those from the play-based kindergartens on every measure that was used. In particular, they were less advanced in reading and mathematics and less well adjusted socially and emotionally.”

In another extensive study of poor children, in Ypsilanti, Mich., young boys and girls who were in academic, instruction-based early education programs were, by age 23, more than twice as likely to have arrest records, less likely to be married and suffering from various types of emotional impairment compared with their peers who attended play-based preschool.

These behaviors (pressing academic work on young children) are a direct result of measuring the wrong thing (test scores). If we measured the right things (social development, curiosity, empathy, imagination and confidence), we would engage in a different set of education behaviors (play, socialization, arts programs, open-ended discovery).

After nearly 20 years of reading, observing, teaching and presiding over a school, I’m convinced that this simple statement — “Measure the wrong things and you’ll get the wrong behaviors” — is at the root of what ails education, from cradle to grave. Measuring the wrong thing (standardized scores of 4th-graders) drives the wrong behaviors (lots of test prep and dull direct instruction). In later school years, measuring the wrong thing (SAT and other standardized test scores, grade point averages, class rank) continues to invite the wrong behaviors (gaming the system, too much unnecessary homework, suppression of curiosity, risk-aversion, high stress, eating disorders).

Measuring the right things is more complicated and less profitable. But if we measured, even if only in our hearts, the things that we should truly value (creativity, joy, physical and emotional health, self-confidence, humor, compassion, integrity, originality, skepticism, critical capacities), we would engage in a very different set of behaviors (reading for pleasure, boisterous discussions, group projects, painting, discovery, daydreaming, recess, music, cooperation rather than competition).

And, as the research in early childhood education makes clear, our children would be better at reading and math, too.



Steve Nelson lives in Sharon and New York City.