The Dark Side Of
Nationwide School Tests
By B.K. Eakman

President Bush's education initiative calls for the testing of every student in the nation, but these 'assessments' in the past involved Big Brother-style psychological profiling.
The proponents of President George W. Bush's education initiative, called "No Child Left Behind," believe that they can make schools accountable to parents as well as taxpayers. The centerpiece of this, as it appears in the amendments to the Elementary and Secondary School Act, still in House-Senate conference as Insight goes to press, is a massive nationwide program designed to test every student in grades three to eight in reading and math. Both House and Senate bills propose some $400 million in federal funds to be sent to the states to devise and administer the tests on a state-by-state basis.
By giving tax money to each state to devise its own tests, supporters hope to mollify conservatives on the one hand, who fear national indoctrination by the U.S. Department of Education, and liberals on the other, who dread the consequences of holding educators personally accountable for whether the children they teach actually learn. The language of the House bill, HR1, for example, states in an unresolved contradiction that each state shall demonstrate that it has adopted "challenging academic standards and challenging academic-achievement standards." In the same breath, the bill says that "a state shall not be required to submit such standards to the Secretary."
The problem is that "academic standards" as defined by common sense and by lawmakers tend to be meaningless when defined by educators. The bill calls for "challenging academic-content standards in academic subjects that specify what children are expected to know and be able to do" and contain "coherent and rigorous content and encourage the teaching of advanced skills." Yet both House and Senate bills shy away from using the term "tests" and substitute the edu-speak word "assessments."
The reason is that public education during the last 30 years has tended against testing for knowledge of content, instead emphasizing a psychological assessment of a child's needs, background and ability to conform to the group. A "test" is an objective measure of a child's ability to solve a problem; an "assessment" is a social scientist's speculation about the environmental conditioning of the child.
Thus the "assessment" of a child's ability to read or to do math in the current testing already in use has more to do with probing the child's psyche and teaching him or her to conform to group values than with testing ability to add two plus two. The leading educational experts will read the bill's language as a license to invade the privacy of every child in the country rather than hold failing schools accountable. And since the bill necessarily honors the principle of local control, it is likely the local educational bureaucracies doing the controlling will welcome the bill as a $400 million slush fund to do exactly what they have been doing to thwart educational reform.
The trouble with school tests begins with the increasing inclusion of sophisticated "behavioral" components that encompass a wide variety of lifestyle and opinion data, nailing down student proclivities, social attitudes and parent-inculcated worldviews. Combined with the plethora of "health" (sex and drug) surveys, mental-health screenings, diary/journal-keeping and other miscellaneous questionnaires - mostly taking place in the classroom under cover of academics - testing has become more equated with personality inventories than proficiency exams. In that context, what passes for testing even may undermine the accountability President Bush advocates.
The case against standardized tests hinges on the quantum leap in data-gathering, cross-matching and information-sharing capabilities, with all the accompanying problems associated with data-trafficking, invasion of privacy and consumer profiling. Barely a week goes by that a publication somewhere doesn't carry a story detailing a new affront to what used to be considered "nobody's business."
One of the earliest examples of psychological data-gathering under the cover of academics occurred in the pivotal 1980s, when enormous breakthroughs in computer technology were being piloted with federal funds in selected localities. One of those was in Allegheny County, Pa., initiated under the eight-state Cooperative Accountability Project. A handful of parents - among them, Gen Yvette Sutton, Anita Hoge and Francine D'Alonzo - got wind of a standardized academic test "no one could possibly study for" being disseminated in the McGuffey School District: the Educational Quality Assessment (EQA). After several unsuccessful attempts to gain access, a trip to the state education agency in Harrisburg finally yielded the facts. Not only did more than one-half the questions not relate to factual knowledge, but numerical codes next to the questions as printed on the administrative version of the test turned out to correlate with specific "remediating" curricula. It included questions such as:
I get upset easily at home:
[a] very true of me;
[b] mostly true of me;
[c] mostly untrue of me;
[d] very untrue of me.
You are asked to dinner at the home of a classmate having a religion much different from yours. In this situation I would feel:
[a] very comfortable;
[b] comfortable;
[c] slightly uncomfortable;
[d] very uncomfortable.
There is a secret club at school called the Midnight Artists. They go out late at night and paint funny sayings and pictures on buildings. I would JOIN THE CLUB when I knew
[a] my best friend had asked me to join;
[b] most of the popular students in school were in the club;
[c] my parents would ground me if they found out I joined."
This last question, in particular, got parents' attention. It presumes that the child will join the club under some circumstances, including the desire to provoke parents. They thought the question more or less asked: "How can we get this kid to vandalize property?"
The EQA had 375 questions covering attitudes, worldviews and opinions - mostly hypothetical situations and self-reports. There were 30 questions on math and another 30 covering verbal analogies - just enough academic questions to appear credible.
Every such test is distributed with professional literature for the educators - which is strictly off-limits to the parents. The EQA told educators it was testing for: the student's "locus of control," his "willingness to receive stimuli," his "amenability to change" and whether he would "conform to group goals." In lay terms, these translate to: Where's the child coming from? Is he easily influenced? Are his views firm or flexible? Is he a team player who will accede to group consensus? Choice "b," then, was the preferred response to the Midnight Artists question because it reflects a willingness to "conform to group goals."
Today, such testing is more sophisticated. A fascinating aspect of a recent Michigan Assessment, for example, was that regardless of the section - reading, science, geography - the questions all sounded like social studies. For example, there was nothing about topography in the geography section; it covered "global issues" - overpopulation, colonial victimization and redistribution of resources to Third World countries. The writing-sample topic? "Coping With Change."
Five science questions for fifth-graders concerned universal child fingerprinting, but involved no science. The multiple choices, even the "incorrect" ones, seemed more like endorsements than questions: "fingerprinting doesn't hurt," "lost children can be identified," etc. Not a single "down side" was offered. The one question that sounded like a question was so simple that one could reasonably have asked whether this were the reading or the science section: "Fingerprinting is MOST useful in which of the following jobs: [a] police work, to help in crime fighting; [b] window washing, to help clean windows; [c] auto mechanics, to help cars run better; [d] teaching, to help kids learn to multiply."
Task I from the history section - on women in combat - was "Interpreting Information." Prefaced in small print was, "Directions: Read the following hypothetical information about a public policy issue. Use it with what you already know to complete the tasks that follow."
Parent activists Deborah DeBacker of Troy, Mich., and Joan Grindel of Bloomfield, Mich., say it's doubtful fifth-graders either understood or acted upon the term "hypothetical." In any case, the only interpretation one could draw from the data provided is that women should be in combat. Despite assurances in the essay instructions that the student's views per se don't matter, it's clear that any view not supported by those "hypothetical facts" in the data section will be judged insufficient to warrant a top grade. In the example, testers actually begin the paragraph for the pupil: "I think that women members of the military should definitely be allowed to participate ."
Questionnaires, curricula and activities that target the belief system are called "affective devices." Psychology texts describe the belief system as made up of attitudes, values and worldviews existing below the level of conscious awareness. Affective means "noncognitive," "dealing with emotions and feelings" rather than the intellect. Using affective-questioning techniques makes it easier to test the subject's belief system. Some go so far as to test for "psychological threshold." The teacher's guide to Pennsylvania's 1986 citizenship curriculum defined this threshold as "the severity of stimulus tolerated before a change of behavior occurs." The manual explained that "it is possible to assess not only the students' predisposition [toward certain reactions] but also to provide some measure of the intensity of that predisposition across a wide spectrum of situations."
Some profiling instruments are explicit and blatant, such as Pennsylvania's and Michigan's, while others are more subtle. Most states label them "assessments" rather than "tests," further confusing the issue for parents. Regardless of the label, opponents claim that personality testing in the context of an academic setting, and the psychotherapeutic sales packages (curricula) that typically ensue, portend a high-tech threat not only to privacy but to a child's future employability and freedom of conscience.
Then there are the student-identification methods applied to "confidential" tests and surveys the testers say are not "individually identifiable." This doesn't mean, however, that students are not "individually identified." Confused? The National Center for Education Statistics 1993 Field Restricted Use Data Procedures Manual explains this semantic sleight of hand. Techniques range from simple bar-coding and "slugging" to more-complicated exercises such as "sticky-labeling" and inserting "embedded identifiers."
To the testers, however, the term "confidential" means "need to know." The "confidential" label casually applied by officials to modern testing and survey devices invariably is taken for anonymity, thereby masking the fact that: (l) higher scores are accorded "preferred" viewpoints, (2) curriculum is modified and targeted to specific groups of children to correct "inappropriate" attitudes and, more ominously, (3) certain views that once were considered "principled" now are deemed "rigid" and associated with mental illness or psychological defects.
Among the at-risk "indicators" are viewpoints and behaviors deemed by the testers to be what they call "indicative of a rigid or underdeveloped belief system." Pupils are referred to psychologists for "remediation" to render their attitudes and responses more "realistic." Several professional papers, beginning with the acclaimed 1969 Behavioral Science Teacher Education Project (BSTEP), place "firm religious belief" in the "rigid/inflexible" category. BSTEP also projected a world "so saturated with ideas and information [by the 1990s that] few will be able to maintain control over their opinions."
So far is all this testing and evaluation from confidential that today's burgeoning computer cross-matching capability of public and private records has launched an information industry of data traffickers and information brokers. Some are licit and others black-market, but they cater to the needs of employers, credit bureaus, universities, corporate spies and government agencies.
Of course, evidence of serious peril to our American presumption of "personal affairs" was being debated among high-ranking educators as far back as 1969, when Wolcott Beatty wrote his seminal work, Improving Educational Assessment and an Inventory of Measures of Affective Behavior. Dozens of related publications followed, documenting a slippery slope from conceptual design of a test that would evaluate and compare effectiveness of learning programs to a federal-funding carrot that would ensure massive personal-data collection with automatic-transfer capability to federal and international databases.
In 1970, L.J. Chronbach's Essentials of Scientific Testing sounded the first alarm: "Coding of records is not a full safeguard. Identity can be detected by matching facts from the coded questionnaire with other facts that are openly recorded."
By that time Dustin Heuston of the renowned World Institute of Computer-Assisted Technology (WICAT) in Utah uttered his prophetic assertion: "We've been staggered by realizing that the computer has the capability to act as if it were 10 of the top psychologists working with one student. Won't it be wonderful when no one can get between that child and that curriculum?" Behavioral-science gurus Richard Wolf (Teachers College, Columbia University) and his colleague, Ralph Tyler, openly were advocating a need for surreptitious methods of data collection and student identification as early as 1974 in their coedited book, Crucial Issues in Testing. They called for unified coding and standardized definitions to enhance cross-matching and data-sharing - from elementary schools on into the workplace.
Wolf supported "the permissibility of deception" in school-testing based on "the rights of an institution to obtain information necessary to achieve its goals." He stated that, danger or not, there "are occasions in which the test constructor [finds it necessary] to outwit the subject so that he cannot guess what information he is revealing. From the constructor's point of view this is necessary since he wishes to ascertain information that the individual might not furnish if it were sought directly. A number of personality tests fall into this category."
Despite admonitions, the lure of computerized cross-matching proved too enticing. In 1981, the first education databanks were launched: the Common Core of Data, the Universe Files and the Longitudinal Studies. In what is perhaps the most evidential document on the subject, "Measuring the Quality of Education" by Willard Wirtz and Archie LaPointe, the writers outline the U.S. Education Department's (ED's) intention to ignore the legal and ethical warnings against privacy invasion:
"Getting into the students' personal characteristics and situations invariably prompts warnings that the NAEP [National Assessment of Educational Programs] purpose is not to analyze human development, and injunctions against confusing the measurement of educational results (outcomes) and the analysis of cause (inputs). But it is being recognized increasingly that the measurement of achievement is incomplete without the accompanying identification of whatever educational circumstances may affect these results."
More prophetically, Wirtz and LaPointe wrote: "A different kind of assessment would help correct the tilt in the educational-standards concept toward functional literacy and away from excellence."
Direct education away from excellence? That's right. The authors detailed how a clearinghouse-style database incorporating demographic and psychological-profiling data would help steer schools toward what these "experts" deemed a more realistic ideal: mere functional literacy.
Policymakers at the ED quickly moved to shelve concerns about student and family privacy. For example, James P. Shaver wrote a detailed monograph, National Assessment of Values and Attitudes for Social Studies, published through the Office of Educational Research and Instruction (OERI), a division of the U.S. Department of Education. But by then there was no need to hide intent because OERI already had brought in four computer experts from Utah's WICAT to prepare a working paper for the first consolidated education database.
In 1986, "A Plan for the Redesign of the Elementary and Secondary Data Collection Program" was finalized, incorporating attitudinal, lifestyle and value information. It fell to the federally funded Council of Chief State School Officers (CCSSO) to ensure state/federal compatibility of computer systems and promote collection of data at the local level. In a 1985 speech, CCSSO Director Ramsey Seldon placed "coordination of educational assessment and evaluation" on the highest priority, promoting the exchange of information about private citizens and their children in the name of comparing educational achievement.
Today, the three original education databases are part of a mammoth data-tracking/sharing system called the SPEEDE/ExPRESS. Among other capabilities, data can be transmitted to universities and prospective employers via WORKLINK, a system set up by the Educational Testing Service.
In 1988, the National Center for Education Statistics named 29 organizations, some with no clear ties to education, that were given automatic access to national assessment data -among them the Census Bureau, the office of the Montana State Attorney General, the Rand Corp. and the Economic Policy Institute. Then technology took another quantum leap - more storage capability in less space, ultrasophisticated search engines, intricate cross-matching methods.
And critics of all this are saying that puts President Bush's national-testing initiative in a different light. And it cuts left and right. After all, if one faction can target a child's belief system and keep records, so can another.
The basic dilemmas remain: If the use of psychographic instruments is legal and ethical, without informed, written, parental consent; if behavior-modification curricula can be brought into the classroom as legitimate learning material; if teachers, or even bona fide mental-health workers, can use the schools to "treat" youngsters for real or imagined psychological problems - then are schools really educational institutions or day-care clinics?
B.K. Eakman, a former teacher and executive director of the National Education Consortium, is the author of Cloning of the American Mind: Eradicating Morality Through Education.

This Site Served by TheHostPros