Finally, defendants speculate that the poor progress of the City's ninth graders is caused at least in part by students who enter the New York City public school system for the first time in the ninth grade. If such students are ill-prepared when they enter the system, that lack of preparation obviously cannot be blamed on the New York City public schools. This argument has as its premise one that is shared by the court: that education is cumulative, and that a student's performance in high school is affected by what he learned in elementary and middle school. Defendants are correct that the evidence at trial did not include any data on the percentage of dropouts, GED recipients, and local diploma recipients who had not attended public school in New York City prior to ninth grade. However, City public [187 [187 Misc.2d Misc.2d 64] 64] school students' scores on standardized tests in elementary and middle school evince the poor performance of City public schools before high school as well. Accordingly, the slow and incomplete progress of the City's ninth graders as they make their way through high school cannot be attributed solely to new students. Student performance on standardized tests is discussed in the next section. 2 . S t a n d a r d i z e d Te s t s The results of evaluative tests administered by the State and by BOE to elementary and middle school students—"cautiously" analyzed per the Court of Appeals' admonition—are probative evidence of the poor quality of education provided by New York City public schools. The first test given youngsters entering New York City public schools before the third grade is the early childhood learning assessment (ECLAS). The ECLAS is not intended to be an achievement test. It is more of a diagnostic tool to see what children know and where they need help. The results of the ECLAS demonstrate what has been noted above: that students entering New York City public schools lack such skills as knowledge of the alphabet, sound/symbol relationships, familiarity with counting and numbers, and vocabulary and concept development. Achievement tests begin in the third grade. Throughout elementary and middle school students take an array of such tests, some of which are given State-wide and are administered by SED and some of which are City-specific and administered by BOE. Both SED and BOE rely on commercially available tests that have been modified to some extent for New York City. a . S t at e Te s t s New York State gives two sets of standardized tests, the pupil evaluation program (PEP) and the program evaluation test (PET). The PEP measures individual achievement in reading and mathematics. By contrast, the PET is meant to measure the general effectiveness of a school in teaching science and social studies. It is not intended to measure the achievement of individual students. During the period 1990-1998 SED used a commercially prepared test called the degrees of reading power (DRP) for its PEP reading test. Before the 1998-1999 school year, this test was designed to test whether students have achieved an [187 [187 Misc.2d Misc.2d 65] 65] extremely low level of reading skills, measured in reference to a particular score known as the State reference point (SRP). As such, the DRP test was only effective at differentiating students at the lowest end of the reading achievement spectrum, and was designed to identify children in need of remedial help. A similarly low SRP was set for State achievement examinations in mathematics. The State reading and math tests were given in third and sixth grades. SED requires that school districts have at least 90% of their students scoring at or above the SRP in reading and math. For the years that data is available, nearly all school districts outside of New York City have been able to achieve this low threshold of achievement. The City has not. In the years 1990-1991 through 1997-1998, New York City did poorly in bringing its students over the low achievement threshold represented by the SRP. Its students did particularly poorly on the reading PEP. On the third grade reading examination, the range of students scoring over the SRP varied from 59% to 69%. Sixth graders scoring above the SRP for reading ranged from 63% to 74%. In math, 92% of third graders scored above the SRP in 1996-1997. In other years, New York City's third graders ranged from 81% to 89% above the math SRP, with a general upward trend from 1994-1995 through 1997-1998. City sixth graders scored at 91% above the math SRP in 1997-1998. In other years the range was 79% to 88% above the SRP, with a general upward trend in the later years. These percentage scores are an average for the entire City and thus include extremely poor performances in numerous districts within the City. In 1998-1999, concerned about the poor evaluative properties of the DRP, SED switched to a new more rigorous commercial test that measures student performance across a wider spectrum. The test was keyed to the Regents' new learning standards. Instead of a single criterion (the SRP), student achievement was measured by reference to four stepped criteria, with level one being the lowest and level four the highest. Students in levels one and two are characterized by requiring extra help to meet the Regents' new academic standards. The examinations were given to fourth and eighth graders. In reading tests given in 1998-1999, 21.3% of City fourth graders scored in level one, the lowest level, while only 5.8% of fourth grade students in the rest of the State scored in level one. In the same year 17% of eighth graders scored in level one for reading, compared with 5% in the rest of the State. Similar results [187 [187 Misc.2d Misc.2d 66] 66] were recorded for math tests with City's eighth graders performing particularly poorly. On none of the tests did more than half of the City's students score in levels three and four, the higher levels that demonstrate that students are on track to meet the Regents' standards. School performance on SED's PET tests, which concern science and social studies, was similarly low. In the period 1990-1996, the City's average scores were in the lowest 25% for the science examinations and never higher than the 16th percentile for the social studies tests. b . C i t y Te s t s Since 1969 BOE has given standardized tests of its own in reading and math to students in grades 3 through 8, separate and apart from the State's PEP and PET tests. Beginning in 1998-1999 the City adopted the commercial examinations used by SED with some modifications. That same year, BOE