The Myth of Achievement Tests shows that achievement tests like the GED fail to measure important life skills. James J. Heckman, John Eric Humphries, Tim Kautz, and a group of scholars offer an in-depth exploration of how the GED came to be used throughout the United States and why our reliance on it is dangerous. Drawing on decades of research, the authors show that, while GED recipients score as well on achievement tests as high school graduates who do not enroll in college, high school graduates vastly outperform GED recipients in terms of their earnings, employment opportunities, educational attainment, and health. The authors show that the differences in success between GED recipients and high school graduates are driven by character skills. Achievement tests like the GED do not adequately capture character skills like conscientiousness, perseverance, sociability, and curiosity. These skills are important in predicting a variety of life outcomes. They can be measured, and they can be taught.
Using the GED as a case study, the authors explore what achievement tests miss and show the dangers of an educational system based on them. They call for a return to an emphasis on character in our schools, our systems of accountability, and our national dialogue.
Eric Grodsky, University of Wisconsin–Madison
Andrew Halpern-Manners, Indiana University Bloomington
Paul A. LaFontaine, Federal Communications Commission
Janice H. Laurence, Temple University
Lois M. Quinn, University of Wisconsin–Milwaukee
Pedro L. Rodríguez, Institute of Advanced Studies in Administration
John Robert Warren, University of Minnesota, Twin Cities
|Publisher:||University of Chicago Press|
|Sold by:||Barnes & Noble|
|File size:||25 MB|
|Note:||This product may take a few minutes to download.|
About the Author
Read an Excerpt
The Myth of Achievement Tests
THE GED AND THE ROLE OF CHARACTER IN AMERICAN LIFE
By James J. Heckman, JOHN ERIC HUMPHRIES, TIM KAUTZ
THE UNIVERSITY OF CHICAGO PRESSCopyright © 2014 The University of Chicago
All rights reserved.
ACHIEVEMENT TESTS AND THE ROLE OF CHARACTER IN AMERICAN LIFE
JAMES J. HECKMAN AND TIM KAUTZ
Modern societies rely on written tests. Achievement tests—multiple-choice exams that attempt to mea sure acquired knowledge—have come to play an especially prominent role. They are used to sift and sort people, to evaluate schools, and to assess the performance of entire nations. The No Child Left Behind Act requires that public schools administer achievement tests and that the test results influence local school policy.
Achievement tests were created in the mid-twentieth century. Their validity in predicting success in outcomes that matter is not well established. Achievement tests were developed as a way to mea sure "general knowledge" that would be useful inside and outside of the classroom. Their developers claimed to have designed pencil-and-paper tests that would predict success in the labor market, in education, and in many other aspects of life. Because achievement tests have been validated by testing experts, most people assume that the tests accomplish these goals. However, achievement tests are typically validated in a circular fashion using IQ tests and grades, and not in terms of their ability to predict important life outcomes. Some have recognized this circularity and have argued that achievement tests miss important skills. There is scant evidence on what skills these tests miss.
This book evaluates the predictive power of achievement tests for life outcomes by examining one widely used achievement test, the General Educational Development test (GED). The GED test is based on the first modern achievement test. The test is a seven-and-a-half hour exam that claims to mea sure the knowledge acquired in completing high school. It embodies the logic of achievement tests. The GED allows high school dropouts to certify high school equivalency to employers and colleges. Currently, the GED program produces roughly 12% of all high school credentials issued in the United States every year.
On the surface, the GED exam achieves its goal. As mea sured by scores on a variety of other achievement tests, GED recipients are as smart as high school graduates who do not attend college. But passing a test does not, by itself, prove anything. How do GED recipients compare to high school graduates in terms of meaningful outcomes?
On outcomes that matter, as a group, GED recipients are not equivalent to high school graduates. High school graduates outperform GED recipients in terms of their earnings, employment, wages, labor market participation, self-reported health, and college completion. Graduates are less likely to use alcohol, commit crime, or go on welfare.
On average, GED recipients perform somewhat better than other dropouts on most outcomes. GED recipients, however, are smarter than other dropouts even before earning their GEDs. After accounting for their greater cognitive ability, as a group, GED recipients are equivalent to other dropouts on almost all outcomes. High school graduates who obtain their credentials through seat time and hard work outperform both GED recipients and uncertified dropouts.
The GED might be a signal that indicates the greater cognitive ability of most recipients compared to dropouts. However, we establish that the GED certificate provides little signaling value in the market. GED recipients earn the same wages before and after they certify.
Our evaluation of the GED provides strong evidence about the predictive power of achievement tests for outcomes that matter. Cognitive ability—as mea sured by achievement tests—explains the average difference in outcomes between dropouts and GED recipients. Something not captured by achievement tests explains the difference between GED recipients and high school graduates. What is the "dark matter" that the test misses?
We show that achievement tests like the GED do not adequately capture character skills such as conscientiousness, perseverance, sociability, and curiosity, which are valued in the labor market, in school, and in many other domains. Until recently these skills have largely been ignored. However, in recent research economists and psychologists have constructed measures of these skills and provide evidence that they are stable across situations and predict meaningful life outcomes.
As a group, GED recipients lack character skills compared to high school graduates. In adolescence, these deficits lead to higher rates of drinking, drug use, violent crime, truancy, vandalism, early sexual activity, and smoking.
There are a few apparent exceptions to this rule. For some, the GED appears to offer benefits. As a group, women who drop out of high school due to pregnancy and who later GED certify have levels of character skills much more like those of high school graduates than other GED recipients. This group of GED recipients appears to perform moderately better than other dropouts in the labor market, although the differences come primarily from their greater labor force participation. The evidence of any causal effect of the GED for this group is ambiguous. Many GED recipients earnestly seek to turn their lives around. For most, preparation for the GED exam does not compensate for the skills they lack.
Differences in character skills emerge early between GEDs and high school graduates. Even at age six eventual GED recipients tend to be relatively smart but exhibit behavioral problems. These findings suggest that many young children are destined to drop out of high school, a view shared by many social scientists.
A prime example of a study claiming early life determinism is the influential and inflammatory book by psychologist Richard Herrnstein and po liti cal scientist Charles Murray, The Bell Curve (1994). Herrnstein and Murray made a major contribution to psychology and social policy by conducting one of the first studies to use meaningful life outcomes in assessing the predictive validity of an achievement test—in their case, the Armed Forces Qualification Test (AFQT). They find that AFQT scores weakly predict success in a variety of life outcomes. However, they do not stop there. They claim that the AFQT test measures the same dimensions of cognition as IQ and that IQ is highly heritable. In their dystopic vision of American society, public policy cannot influence the skills that affect success in life. Like most people, Herrnstein and Murray overlook character as an important predictor of success and as an avenue for social progress, and also ignore the evidence on the malleability of IQ.
Investment and interventions can foster character. The Perry Preschool program is a telling example. Young (age 3–4) low-IQ African American children were given early stimulation. Participants were taught how to plan, execute, and review tasks. They learned to work with others when problems arose. The program was evaluated by the method of random assignment, and participants and controls have been followed through age 40. The program had no long-term effect on IQ scores. By the Herrnstein and Murray criteria, it failed. Nevertheless, the program improved outcomes for both boys and girls, yielding a rate of return that outperforms the stock market in typical years. Heckman, Pinto, and Savelyev (2013) show that the program worked by improving character.
Because both cognition and character can be shaped, and change over the life cycle, we refer to them as "skills" throughout this book. An older terminology refers to them as "traits," conveying a sense of immutability or permanence, possibly due to their heritable nature. The literature surveyed in Chapter 9 shows how these "traits" can be enhanced. Our distinction between skills and traits is not just a matter of semantics. It suggests new and productive avenues for public policy.
Character training is not a new idea. Aristotle mentions it in the Nichomachean Ethics. Prominent American educators since Horace Mann have noted that successful schools produce more than problem-solving skills and factual knowledge. Schools also mold character.
Recently, many have come to view character education as the sole province of the family or the church. Families are important producers of both cognition and character. However, the American family is under severe challenge. Single-parent families—which provide fewer resources for development of character and necessary life skills—have become pervasive. Even many intact families are stressed because of diminished resources.
This book shows that, as a group, GED recipients lack character skills in part owing to their relatively disadvantaged family backgrounds. Compared to high school graduates, GED recipients are more likely to come from broken families with low incomes and have parents who invest less in their character and cognitive development.
At a time when many families could use more support, character education has been phased out of schools and the ability of schools to enforce discipline has been weakened. In the nineteenth century, character education was prominent in American schools. They had strict disciplinary standards and taught character directly, often through religious texts. A Protestant vision of morality and character was incorporated into public education.
Five primary forces led to the decline of character education in public schools in the last century. First, the rise of cognitive psychology shifted the focus of American education toward cognitive training and measurement (Bruner, 1956). Second, growing support for the separation of church and state removed religious teaching from the classroom and any forms of moral education or character education that smacked of religious training. Third, the "legalization" of schools increased the rights of students but reduced the autonomy of teachers and the use of disciplinary measures that could be used to instill character (Arum, 2005). Fourth, cultural relativism became more widespread in society. The notion of a core set of character skills that was universally agreed upon fell out of favor. Those advocating a core set of values and evaluation of character were accused of seeking to impose their (middle-class) values on others. Fifth, the research of Walter Mischel (1968) appeared to establish that there are no stable personality skills. If character was ephemeral, there was no point in measuring it or trying to foster it. These trends contributed to the demise of character education in schools, which in turn exacerbated the problems created by the emergence of single-parent families in shaping the character of youth.
Character education does not necessarily infringe on the liberties of students or families. Character education has moral components, which some conflate with religious values. Character skills are universally valued regardless of any religious orientation, although churches, temples, and mosques produce character. Removing religion from schools does not require removing character education from the curriculum or preventing evaluation of the character of students. Virtually all parents want their children to be hardworking, honest, per sis tent, creative, curious, self-controlled, and excited by learning. Curricula that teach these skills in conjunction with cognition are promising ways to foster successful lives while maintaining the sanctity of the family and preserving the separation of church and state.
The curriculum in schools backed off from evaluating and fostering character to focus primarily on producing and measuring cognitive development. Belief in the predictive value of achievement tests became pervasive. It led many to view the GED certificate as equivalent to a high school degree. In some states, the GED is legally mandated to be equivalent to a high school degree for the purposes of employment and admission to postsecondary education.
This book shows that this faith in tests deceives students and policy makers and conceals major social problems. The GED misleads students when they are making educational decisions. High school students as young as sixteen can take the GED. Adolescents are impressionable, and for many the GED seems like an attractive alternative to finishing school. We show that having a GED option available induces students to drop out of high school.
After the GED was introduced in California in 1974, the high school dropout rate increased by three percentage points. More recently, Oregon introduced the GED Option Program in high schools. These programs teach the GED and make it easier for students to GED certify. The Oregon program increased the high school dropout rate in the districts where it was implemented by four percentage points.
The GED deceives its recipients into believing that they are prepared for college. About 40% of GED recipients attend college. About half drop out in the first year. Far fewer complete any degree, but pay costly tuition and forego substantial earnings in quest of degrees that they do not obtain.
The deception runs deep. All GED recipients are not alike. Some are hardworking and acquire skills by preparing for the exam. Despite their hard work and high character skills, these GED recipients are lumped into the same category as the relatively smart but unmotivated students who pass the exam. Employers and colleges might overlook the true achievers among the mass of GED test takers because the GED exam does not discriminate between the motivated and the accomplished and those who just pass its minimal standards.
The GED distorts social statistics and masks in equality. Many social statistics classify GED recipients as high school graduates. This misclassification conceals black–white gaps in educational attainment. If GED recipients are counted as high school graduates, the black–white gap in high school graduation rates has closed substantially. If GED recipients are counted as dropouts, there has been no progress in the black–white high school graduation rate in the last 50 years. Many black GED recipients earn their certificate through remediation programs in jail.
Based on the belief that the GED is equivalent to a high school degree, government programs have channeled substantial resources into producing GED certificates. These resources could have been spent on more effective policies. The success of many adolescent intervention programs such as the Job Corps is judged by the number of GED recipients they produce. This practice distorts funding choices. Government support also helps to explain why the GED program has become so prevalent even though it offers few benefits to most recipients. The cheap fix has become the byword of American public policy. While the direct costs of the GED program are low, it fixes few problems for most GED recipients and creates a host of new ones.
To address problems with the test, the GED testing ser vice is planning to increase its passing standards. This proposal ignores the fundamental problems with the GED, which will not be solved by raising its passing standards. The GED program is a symptom of the deeper problem that American society is failing to produce essential character skills. It is possible to tackle this problem, but not simply by raising standards on achievement tests.
1.2 The Origins of Achievement Testing
The confluence of four cultural and intellectual currents produced the GED testing program and America's heavy reliance on achievement tests. First, technological developments made it cheap to implement multiple choice tests on a large scale. Second, cognitive psychology fostered the belief that cognition is the primary skill required for success in life. Third, for reasons discussed in the previous section, character education and the evaluation of character skills were slowly phased out of schools, partly accelerated by the federal government's entry into public education. Fourth, the accountability movement in government mimicked the logic of private market cost-benefit analysis by using test scores to evaluate and assess a myriad of government programs designed to enhance skills. We now elaborate on these points.
The modern thrust for accountability in schools arose in the nineteenth-century educational reform movements. In the early nineteenth century, Horace Mann introduced the first standardized test used in American schools. The test was an early attempt to evaluate schools by their output—the knowledge they produced—rather than by their inputs. The instrument he devised was very crude. As noted in the Preface to this book, Mann saw the limitations of his primitive achievement test (Mann, 1838). However, Mann's test was not widely implemented because grading it was laborious and time intensive. It would be another century before his ideas for standardized testing became prevalent.
In the absence of reliable output-based measures, nineteenth-century educators largely evaluated schools using input-based measures (e.g., standardized curricula). The input-based system was criticized. Teaching was often rote-based. Many critics commented unfavorably on the rigid disciplinary environments in these schools, which were intended, in part, to instill valued character skills in students.
Excerpted from The Myth of Achievement Tests by James J. Heckman, JOHN ERIC HUMPHRIES, TIM KAUTZ. Copyright © 2014 The University of Chicago. Excerpted by permission of THE UNIVERSITY OF CHICAGO PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of ContentsPreface
Part I. Introduction
1 Achievement Tests and the Role of Character in American Life
James J. Heckman and Tim Kautz
Part II. The History of the GED
2 An Institutional History of the GED
Lois M. Quinn
3 Growth in GED Testing
John Eric Humphries
Part III. Evaluating the Benefits
4 Who Are the GEDs?
James J. Heckman, John Eric Humphries, and Tim Kautz
5 The Economic and Social Benefits of GED Certification
James J. Heckman, John Eric Humphries, and Tim Kautz
6 The Military Performance of GED Holders
Janice H. Laurence
Part IV. The GED Creates Problems
7 The GED Testing Program Induces Students to Drop Out
James J. Heckman, John Eric Humphries, Paul A. LaFontaine, and Pedro L. Rodríguez
8 High- Stakes Testing and the Rise of the GED
Andrew Halpern- Manners, John Robert Warren, and Eric Grodsky
Part V. What Can Be Done to Promote Character?
9 Fostering and Mea sur ing Skills: Interventions That Improve Character and Cognition
James J. Heckman and Tim Kautz
10 What Should Be Done?
James J. Heckman, John Eric Humphries, and Tim Kautz
List of Contributors