Reviewer: Prof. Wayne Bishop, Math & Comp Sci, Cal State LA (213)343-2159

 Dear Prof. Tsang                                                                       December 23, 1997

You sent me an paper entitled "Effects of the Connected Mathematics Project on Student Attainment" and asked my opinion.  Glad to oblige.

The Balanced Assessment (BA) by its very name indicts the ITBS as an "unbalanced assessment" consistent with the prejudices of the current reform movement in mathematics.  When BA has been studied carefully for a decade or so to see long term correlation with stronger performance, I'll be ready to give it some credence.  To this point it should be marked "under construction" and ignored.

The ITBS is almost the diametric opposite, I trust it implicitly and wish our state would have chosen it as its new assessment instrument instead of the Stanford-9.  It is the researchers themselves that I don't trust. They imply that these are balanced groups of students, a fact clearly not valid since there is more than a year's difference in grade level equivalent scores for classes being compared at grade 8.  That in itself could explain the reported difference at the end.  For "size of n" this small, very careful statistical analysis must be done to conclude anything useful at all.  If n were classes, instead of students, and nearly imperceptible beginning values, the same results would have to be considered meaningful.  These aren't.

Perhaps as indicative as anything of the uselessness of the work is the results of the non-CMP groups at grades 6, 7,  and 8.  Either the groups are outrageously different student groups or the curricula they use is hardly "traditional".  From Grade 6 to Grade 8, they declined 0.7 of a year's growth in Problem Solving?  and no growth overall, a respectable 8.6 grade level in spring of 6th to a less respectable 8.6 in spring of 8th!   As was commented by the authors in paragraph 3  of page 8, all of the scores decreased markedly when computation was included, roughly a whole grade level year of the CMP and a half of a year for the non-CMP students across all grade levels.  The authors imply a problem with the ITBS norming sample.  That is ridiculous, you can bet the house on it.  The CMP students are substantially worse at developing computational competence in an educational environment where the students to which they are being compared are already terrible.

Indicative is the presence of a University of Pittsburgh "researcher" and QUASAR connection (see footnote 3 of page 1).  Here is an excerpt of a paper that documents what QUASAR did to algebra readiness in its single California pilot. That pilot was praised to the nation as highly successful by the then president of the NCTM.  Not surprisingly, the NCTM has refused to publish this information because of a mind-set similar to the authors of this "study" of CMP performance.

Thanks for asking,
Wayne Bishop,
Math & Comp Sci,
Cal State LA
(213)343-2159

Table 1 records the last year of the former program and the first through the third years of the program at the school to which Dr. Price referred and includes the most recent Algebra data provided by the district.  In this table, the A and B grades are grouped, as are the D and F grades. Course grades can be suspect, of course, and a good, objective measure of Algebra success is available to California high schools through the Mathematics Diagnostic Testing Program (MDTP).  Unfortunately, MDTP results were not made available.

Table 1 Summarizing, three years into program the number of students willing to elect algebra in eighth grade at the school had dropped by a third and their success rate had gone from so-so to abysmal.

Eighth Grade Algebra        1990 1991 1992 1993

                  A&B  Grades   37%  19%   5%  11%
                        C  Grades   36%  33%  27%  16%
                   D&F Grades   27%  48%  68%  73%
           Number Enrolled    56      58      37     37