The Politic - National
Issue: 2/21/05

Educational Testing Policy
By Richard P. Phelps

The U.S. public has consistently favored the use of standardized testing in the schools, preferably with consequences (or "stakes") riding on the results, since first polled on the topic several decades ago. Depending on how the question is framed, those in favor of high-stakes standardized testing outnumber those opposed from two-to-one to twelve-to-one. Parents are stronger supporters of high-stakes testing than non-parents, and that support does not budge when offered the possibility of their own progeny failing.

Results from different polls approaching the topic in different ways suggest that most Americans would like to see high-stakes tests administered at least once every grade level. The typical U.S. school district, however, offers just one or two high-stakes standardized tests in twelve years of elementary and secondary school. Few public programs attract such a high level of support. Likewise, few public programs are afforded less serious consideration in policy discussions.

Mind you, I am not suggesting that policy makers ignore the issue completely, nor that testing receives no political support. Quite obviously, large-scale testing programs do exist in the United States, many politicians profess support for them, and policy discussions are engaged.

With only a few exceptions, however, U.S. educational testing programs fall far short of what the U.S. public wants, and far short of what most industrialized countries have. And, unless the politics of the issue changes dramatically, in some manner I cannot foresee, it is likely to remain this way.



Comprehensive Testing Systems are Multi-leveled and Multi-targeted

A comprehensive testing system is one that captures all the benefits standardized testing offers, and does it for all students, not just some. Large-scale, high-stakes educational tests offer three benefits:

1. Information that can be used for diagnosis (e.g., of individual students or teachers, of schools, of school programs);

2. Efficiencies from alignment, when the tests are matched to curricular standards and teachers teach to those standards (and, yes, teach "to the test," as they are supposed to do with standards-based tests); and

3. Motivation to study and to attain goals.

The best testing regimes, such as one finds in many European and Asian countries, tap all three sources of benefits through multi-level and multi-target systems. "Multi-level" means that high-stakes tests are administered at more than just one educational level. Typically in European and Asian systems, students face high-stakes tests at the beginning and/or end of more than one educational level (e.g., at the end of primary school, the beginning and/or end of lower secondary school, the beginning and/or end of upper secondary school, and the beginning and/or end of postsecondary education).

"Multi-target" means that every student, no matter where they are in the range of achievement or in their choice of curriculum, faces a high-stakes test that, ideally, offers a challenging, but attainable, goal. In some systems, tests are set at multiple levels of difficulty, and offer multiple levels of certification (e.g., a "regular" diploma and an "honors" diploma). In other systems, different tests cover different subject matter (e.g., general, vocational, or academic; literature, math & science, technology, or social science).

In the United States, high-stakes tests are uncommon at any but the upper secondary level. Moreover, with very few exceptions, they are single-target tests-each and every student, no matter what their level of achievement or ability, course selection, or curricular preference, must meet only one common standard of performance.

Ironically, largely socialist Europe, with its relatively small socioeconomic and academic achievement disparity, acknowledges that children are different and offers them a range of academic options and multiple achievement targets. In the more libertarian United States, with its relatively large socioeconomic and academic achievement disparity, pressure is brought to bear for all children to take the same curriculum (i.e., what is often called the "college track") and a single academic achievement target is set for all.

When only one academic achievement target is offered, by necessity it must be a low target. If it is not, huge numbers of students can fail and the educational system can collapse on itself. When the single target is low, responsive school systems focus effort and resources toward bringing the lowest-achieving students up to that target. Unfortunately, they also may neglect the average- and higher-achieving students or, in the most perverse cases, deliberately hold them back.



The No Child Left Behind Act

The federal No Child Left Behind (NCLB) Act set in place what is largely a testing program. NCLB, however, falls far short of a comprehensive multi-level, multi-target high-stakes testing system. State NCLB testing systems typically set only one target (for schools), no stakes for students (so little motivation to take the test seriously), and curricular alignment can be less than perfect. Yet, as little as it may be, NCLB is commonly characterized by educators and journalists as either being too much or else marking the limit of what our schools can bear.

NCLB is modeled after a Texas testing program supported in the 1990s by then-Governor George W. Bush. The general outline of the program had been initiated in the 1980s by one Democratic Texas governor and then developed during the administration of another. But presidential candidate Bush championed the program as his own, and so did his critics.

During the presidential campaign of 2000, a deluge of negative press coverage swamped the media. I conducted a tally from spring 2000 into the fall. Articles on standardized testing, most of them about Texas, ran at an average rate of about six a day. Most of them were uniformly negative about tests. In addition, half a dozen anti-testing books were released during the year, with their authors interviewed frequently in the media.

Dozens of professional education groups, such as the American Association of School Administrators, the National Council of Teachers of English, the Association for Curriculum Supervision and Development, vehemently denounced standardized testing programs in general and candidate Bush's proposal in particular. Surf the web sites of these groups even today and one will find an unadulterated diet of one-sided anti-testing research, information, and sloganeering. Such groups form the Democratic Party base on education issues.



Base-ically Opposed

I relished the prospect of standardized testing being a key issue in a presidential campaign. In normal times, public debate on the issue is dominated by testing opponents. The year 2000 held promise for a wider, more balanced discussion. To encourage it, I conducted some research myself, the results of which contradicted the most prominent accusations. I sent the work to journalists and to any Republican group I could find, assuming that they would want to defend their standard-bearer. I informed both groups of the aforementioned negative press deluge and encouraged efforts to tell the other side of the story.

Hundreds of letters, phone calls, and email messages later, I was stunned. Hardly any of them cared. Education journalists seemed quite content to tell just one side of the story. Most of the Republican groups showed no interest in the issue and the few that were interested were strongly opposed to standardized testing as a big government intrusion into local and family affairs. Not until half a year after George W. Bush was elected president did any of the Republican groups put any substantial effort into standardized testing advocacy. The base of each of the two major political parties seemed to dislike educational testing just as much as the overwhelming majority of the U.S. public liked it.



Republicans Do Not Necessarily Represent the Other Side

Yet, one finds only representatives of the two party bases in most media coverage of the topic. Many journalists assume that the Republicans must represent "the other side" of the testing issue and dozens of them call one of just a handful of Republican policy brokers for quotes or references to good sources. That most education journalists call the same few people every time for their stories suggests how uniform pack education journalism can be. That most education journalists assume the other side of a story is represented best by professional politicos demonstrates what little credence they give to education research as an objective, scientific pursuit.

Republican education policy development is run, more or less, like the petty fiefdom of overly acquisitive local labor bosses-all functions being staffed by the personal favorites of just two or three individuals. Look at the education policy work conducted either by Republican think tanks or promoted by Republican-oriented education advocacy groups and odds are it was produced by: (1) a certain former Assistant Secretary of Education-now a foundation director-or one of his dozens of former staffers; (2) a certain Ivy League Political Science professor or one of his former students or staffers; or (3) one from a group of a dozen or so sympathetic academic economists.

To counter the vast, experienced, and broadly-knowledgeable armada of talent advising education's vested interests and allied with the Democratic Party, the Republicans launch a bathtub boat flotilla, manned by a small inbred crew. The Republican thinkers do read and acknowledge each other's work, and, more broadly, some of the little work done on the topic by economists. But they declare nonexistent an enormous research literature accumulated by education planners and practitioners, program evaluators, and psychologists. Given that educational testing was invented by psychologists, that all tests are developed by psychometricians, that tests are administered by educational practitioners, and that testing programs are evaluated by program evaluators, the Republicans' focus seems a wee bit narrow. Most GOP groups dutifully disseminate the work emanating from this little group and, just as dutifully, ignore the vast majority of education research and information.

By the way, you can forget about finding pro-testing advocates among the fringe political parties. Ralph Nader is responsible for founding the most extreme anti-testing group in the United States. Libertarians generally oppose enforced standards and large-scale testing as big government intrusion.

That many politicians support education standards and standardized testing, despite discouragement from their core supporters and close to universal disapproval from journalists, should be considered refreshing to any democrat. Our elected officials are often accused of pandering to the media and to their core supporters while they ignore the wishes of their constituents. Perhaps that is often the case, but it is not on this issue.

The intent of politicians to respond to the public mandate supporting educational testing may be honorable and their actions to implement testing programs in the face of often vitriolic opposition should be considered heroic. But politicians are not psychometricians, as the on-again, off-again pattern and perpetually stunted form of testing programs in the United States affirms.



Educational Testing: An Orphaned Policy

Though they are rarely heard from in policy debates, many of the world's foremost testing experts work in North America. Ironically, they live in what could be considered an educational testing backwater.

There must be thousands of education researchers in the United States who feel as frustrated as I do. Not only is the vast majority of quality research and information on educational testing ignored by journalists and the small group of celebrity researchers they talk to, extraordinarily often it is declared nonexistent. As a result, the American public and the American politician are persistently misinformed, and their desires reliably unrepresented in policy discussions.

What is a political centrist and proponent of a hugely popular public policy to do? Arguably, our country has not seen a viable centrist third party since 1912. But perhaps therein lies the only hope for a political group that will champion the sensible educational testing policy favored by the overwhelming majority of Americans and adopted by most of the planet's industrialized countries. "Nothing can stop a Bull Moose!" Nothing, that is, except a two-party system.


Richard P. Phelps is author of Kill the Messenger: The War on Standardized Testing (Transaction, 2003), and editor of Defending Standardized Testing (Lawrence Erlbaum, January 2005).