choice of standardized testing

["GERALD BRACEY" <gbracey1@verizon.net>]

Sherman:

Blame Bobby Kennedy.

Aside from attacks on the IQ tests, ongoing since the 1920's, the criticisms you refer to never had a lot of currency and virtually no influence on the field of psychometrics.  I headed the testing programs for Virginia starting in 1977 and during my tenure the legislature enacted a raft of testing programs.  There was never any debate over whether or not testing was a valid means of measuring achievement.  People drew on their own experience as test takers--the state superintendent suggested that the cut score on the state's new minimum competency test be set at 60 because when he was in school a 60 earned you the lowest passing grade.

I can offer you a history that I think is true, but have not tried to validate systematically.  

In the 20's and 30's there was a lot of experimentation with different types of items (in his 1929 book on testing, Giles Ruch argued for the abolition of oral classroom tests which tested one child at a time; the SAT contained essay questions until 1941), but the fast and cheap multiple choice item type came to exclude all others, especially after E. F. Linquist developed ways of scoring the tests electronically.  Teachers never accepted these tests as representing what they were trying to accomplish in the classroom.  The psychometricians, however, were perfectly happy because their interest was mostly in discriminating among people to make predictions about future behavior and the tests were designed to do that.  They also thought that their tests were objective.  Ruch's book was titled The Objective or New-Type Examination.  They also thought that testing was an exact science and they knew they were right.

This disconnect had little importance as long as tests had no impact on curriculum and instruction.  But in 1965 came ESEA.  Up to that moment, social programs were generally evaluated in terms of inputs and processes: we had X dollars and we spent them on Y programs.  Bobby Kennedy wanted to know, "and then what happened?"--wanted evidence of program impact and built evaluation into ESEA.  When people looked around for instruments to evaluate with, they found tests.  The evolution of the multiple choice test had wiped out most experimentation with other instruments.

There was also a growing distrust of teachers and a growing interest in tests outside of psychometrics.  In the runup to Sputnik, American school were intensely and extensively criticized.  Educational Wastelands: The Retreat from Learning in Our Public Schools was published in 1953.  National Assessment was created in the late 1960's and its sponsor, Commissioner of Education Frank Keppel, wanted it because he believed it would document low performance by schools.  I left the country in 1973 when everyone was talking Piaget and informal learning.  When I got back in 1977 everyone was talking "back to basics" and using tests to measure "the basics."  

The College Board panel's 1977 report did nothing to "relegitimize" the SAT.  Believers had always believed and critics were yet to be widely heard (although officers at ETS were well aware of criticisms; Nairn and Nader's book would appear in 1980).  "A 14-year decline in the Scholastic Aptitude Test scores is largely a result of an increasing number of minorities, poor, women and academically inferior students' applying for college entrance..." was how Ted Fiske characterized it in the NYTimes.  He left out television which the panel also invoked.

"A Nation At Risk" in 1983 raised the level of distrust of information flowing from schools.  We need something the schools can't screw with.  What have we got?  Tests.  

There were dissenters, of course, notably the NEA and the Huron Institute's National Consortium on Testing headed by Vito Perrone, George Madaus and Walt Haney.  The onset of the minimum competency testing madness which spread to 35 states led to a national conference on testing in 1978.  In an after dinner speech, Claiborne Pell leaned ominously over the podium and asked "How many of you favor a national test?"  Of the 300 attendees, two test publishers and, for some reason, the president of the National Student Association raised their hands.  It also led to a famous 1977 or 1978 debate between Jim Popham and Gene Glass at the annual ECS conference on large scale testing, and to a three-day "trial" of minimum competency testing sponsored by NIE in 1981 with the pro side led by Jim Popham and the con side led by Madaus and Bob Linn (I joked at the time that it was ironic to have the pro side led by a con artist).  

There was no "verdic" although Madau and Linn were overshadowed by Popham's Jimmy Cochran style.  That was left up to the viewers of the three hours of testimony shown over PBS.  MCT died essentially when people realized it was a hoax: cut scores were set so that enough people failed to satisfy most of those who called for the tests in the first place but not enough to cause an uprising of citizens--something NCLB will have to face soon (George C was right when he said the most honest way to build a test is to first decide how many people you want to flunk).  "A Nation At Risk" assisted in the death by saying MCT was not enough, the focus should be on excellence.

Lots of history floating around.  Check it out some time.

Jerry