'Bell curve' excuse can lead to lower expectations: the achievement gap for any individual student need not exist, despite agruments that some groups will always perform poorly

Leadership, Sept-Oct, 2003 by George Manthey

While I was a graduate student, a classmate of mine--an ardent feminist--wrote an extensive term paper on gender differences in the performance of infants. To her dismay, she discovered that there did seem to be some differences in responses to and manipulation of geometric shapes. But, she also discovered a related bit of data tha has proven to be the most memorable a related bit of data that has proven to be the most memorable fact I learned while getting my teaching credential. The range and variation of scores of the infants studied were greater within each sex than they were between the sexes. Therefore, even though the mean scores may have produced some statistically relevant differences, it was impossible to predict how any individual would behave.

Consider the statistical construct of the bell curve. There is always a gap between the curve itself and the X axis of the graph it is drawn upon. Neither end, theoretically, ever reaches the X axis--although they approach it continually. This means that however tall, flat or skewed a bell curve--or whatever the mean is of the data being reported--the two ends of every bell curve are virtually the same. The chances, then, of any individual, regardless of the group he or she may be aligned with (gender, income level, ethnic background) scoring at any particular place within the bell curve are virtually the same. This is true if were are measuring foot size, speed, intellect or scholastic achievement.

Whenever I hear someone share that it is important to have high expectations for student learning for every student group (a la NCLB), a critic invariably points out that the bell curve is just not being taken into account. By doing so it is argued that certain groups, most often the economically disadvantage, will always perform at low levels. But to the best of my knowledge our schools do not teach groups of students; we teach individual students. Success or failure of any individual student cannot be predicted based on whatever group a student might fit within.

In his presentations, Jim Cox sometimes shares the multi-year study of students who grew up in the Bronx which was conducted by a prestigious university. In this study a larger group of students than statistics would have ever projected--based on factors like their income level, the education level of their parents, the propensity of neighborhood gangs, etc.--grew up to be remarkably productive citizens.

The researchers looked further to see if this subgroup had anything in common. Sure enough they did, each had had the same teacher. When this teacher was interviewed in an attempt to discover if and how she had made such a difference, nothing of significance was discover until she made an offhand comment as the interview came to a close: "I loved those boys."

Invariably, when studies are done of students who have become very successful, despite all "the odds," what is discovered is that at least one person in their lives inspired and expected much from them. David Ramirez is currently busier than he might choose to be helping schools implement a mentoring program where adults mentor just one (perhaps more) student who has dropped significantly in achievement levels after having been much higher for two years or more. The results are "API-busting!" These students make remarkable achievement gains. How does Dr. Ramirez suggest they be mentored? By at least weekly check-ins and offers to be of assistance. Very akin to our teacher in the Bronx who "loved her kids."

The achievement gap is very real. So is that gap between the X axis and the bell curve. This second gap demonstrates, statistically, that the achievement gap for any individual need not exist--and cannot be predicted. Imagine if each of California's approximately 30,000 school administrators took on one student who is at risk of failure. If just half of these efforts were successful, 15,000 students who might not have otherwise made it would be on the right track. A very significant step toward closing the achievement gap.

Bell curve results are inevitable whenever the full range of any human performance is measured. However, it's both a statistical and scientifically based fact that bell curve results cannot predict individual performance. Don't ever let anyone claim that they do.

George Manthey is an educational services executive for ACSA.