By John Verzani
On July 21, 2018 the New York Post had an article “Don’t let lowered math standards become a trend.” This in depth reporting was done a few days before the newsroom staff at the Post was cut in half. So let’s ask, do “cuts” have the expected results?
In Fall 2017, CUNY relaxed “cut” scores for the Math Placement Exam from those used in Fall 2016 (though with a different test). The Post reported that “The percentage needing remedial math in particular fell from 73 percent to 52 percent” with the implication that this is a bad thing. Yet, some might argue “too” many students were already in remediation. Let’s look at what happened to the students who would have placed into remediation in Fall 2016 (F16), but who did not place into remediation in Fall 2017 (F17) — the newly non-remedial.
According to the CUNY Student Data Book there were 25,469 first time freshman (FTF) at a CUNY community or comprehensive college in F16 and 25,620 in F17. If we apply the 73% and 52% across the board (which isn’t quite right, as the percent is less at comprehensives) we find that in F16, 6,876 first time freshman would have placed out of “math” remediation and in F17 12,124 first time freshman would have placed out of math remediation.
Data on performance of first time freshman in math at CUNY from F12 through F17 generated by CUNY’s OIRA office was shared with a math discipline working group to study placement by means other than high stakes exams. This data summarized enrollment and pass rate for each CUNY math class with first time freshman. Using this data, we can identify those non-remedial classes at each of the colleges (23 in total) that saw a surge in enrollment of first time freshman between F16 and F17. In F16, these classes had 5,625 students enrolled; in F17 there were 10,183, an additional 4,558 students. These 23 classes account for about 80% of the students who “could” have taken credit-bearing math in the fall (some choose not to, some were in other classes). Analyzing these courses gives us a sense of how these newly placed students did.
First, whatever the outcome, this needs to be recognized as a radical shift in placement and the implications will take a few years to become clear, as departments make adjustments. On the positive side, whereas approximately 3,670 students under study passed a college-credit earning class, in F16 there were now approximately 6,080 who passed. So more opportunity led to a big surge in the number who found success. Placing out of remediation worked well for many students.
However, there was a cost. In F16, 65% of the students (FTF) in these courses passed, in F17 only 59% passed. Ideally — in this writer’s view — the pass rate target for first-time freshman should be 70% at comprehensive and community colleges, and at least 85% at senior colleges (the difference being more withdrawals). This pass rate is dependent on initial placement policies. If it is too high, the placement likely excludes students who could have achieved at that level; if it is too low, the placement likely puts some students into an unwinnable position. For the identified classes, first time freshman are expected to be the highest performing cohort. If fewer than 60% fail to pass, it is likely fewer than 50% of the class, overall, will pass. Such low-performing classes are demoralizing for both students and faculty.
In total, 11 of the 23 classes considered saw pass rates decline by 10% or more; only 1(one) class saw a modest increase; the number with pass rates below 60% climbed from 7 to 10. Therefore, the F17 pass rate is moving in the wrong direction from this writer’s target. More importantly, it moved in ways that depended on the class level.
We don’t have data on these students individually, only aggregated data for first time freshman. However, if we “assume” pass rates are constant for students who would have been non-remedial in either year (they aren’t) and assume enrollments would have been identical (they wouldn’t be) “then” we can “estimate” the difference in pass rate between the cohort of students who would have been non-remedial and the newly non-remedial. That is, for this purpose, the 4,558 additional enrolled students. For example, if the overall pass rate dropped 5 percentage points, and there were twice as many students, then the new students pass rate would be estimated to be 10 percentage points less. If we “assume” all this, then we observe:
* 4 of 5 college algebra classes saw a difference of over 10 percentage points (with the lone one being for a small, special cohort class)
* 4 of 6 intermediate algebra classes saw a difference of over 10 percentage points
* 4 of 6 statistics classes saw a difference of over 10 percentage points
Yet only 3 of 6 Quantitative Reasoning (QR) courses saw a decline.
It would stand to reason, that students who have historically had algebra issues, would find advanced placement in algebra courses more difficult. We can see that the overall pass rate for first time freshman in the College or Intermediate Algebra courses dropped from 64% to 57%, leading to an troubling estimated 45% pass rate in College Algebra and 50% pass rate in Intermediate Algebra for the newly non-remedial.
CUNY, following national trends, does not require intermediate or college algebra skills for its graduates, and campuses have traditionally offered quantitative reasoning courses or elementary statistics courses for its students. While there were declines in pass rates in QR courses, the overall rate is still above 65% for the identified courses and the newly non-remedial seem to have a better chance of success.
In statistics courses the reliance on algebra skills is not so direct, but proficiency of algebra skills appears to be a lurking variable for success in statistics courses. Here we also see troubling indicators: overall pass rates for FTF dropped from 66% to 61%, and the newly non remedial have an estimated pass rate of just 53%,
CUNY recognizes some of this. For the still remedial cohort (though with identified arithmetic skills) CUNY has mandated–over objections of some math faculty–pathways that allow students to earn college credit without having to take elementary algebra. These are typically co-requisite courses. Student success is much better for QR courses with this “pathway”, though some campuses have found success with co-requisite statistics (LGCC is a good example). For the cohort that is newly non-remedial, co-requisite courses are not mandated, though perhaps that extra nudge is still needed for a reasonable chance of success in Elementary Statistics or Intermediate Algebra.
The Post’s implication that CUNY is weakening standards: Clearly this isn’t so! The data suggest math faculty have not lowered their expectations of what passing means, but rather demonstrate the passing rates have predictably gone down. As long as campus provosts support the faculty in assigning the grades as they are earned, a charge of weakening standards is not warranted. Departments then can work on support strategies to build up pass rates in the troubling courses and adjust placement policies as needed.