In her blog post from the beginning of this school year, Suzanne Wells examined a phenomenon in DCPS neighborhood schools on Capitol Hill: the exodus of 4th graders from neighborhood DCPS elementaries to charter middle schools at 5th grade.
As the principals of those neighborhood elementaries attested, the students who left to enroll in charters were performing at proficient or advanced levels.
Now, recent PARCC scores appear to validate those observations city-wide.
At the highest scoring levels in PARCC–reported by OSSE as 4+ (that is, at levels 4 and 5) and 3+ (level 3 and above)–ELA scores declined from 4th to 5th grade within DCPS, while those scores in charters at the same grades went up.
This mirrors a similar shift in PARCC math scores between sectors from 4th to 5th grade:
This shift of the highest test scores between charters and DCPS at those grades (one declining, while the other rises) is mirrored in test score shifts in both sectors from middle to high school, another crucial juncture in DC public education.
That is, while charters appear to have higher PARCC test scores in 8th grade relative to those in DCPS, that changes by 10th grade, such that DCPS appears to have the highest test scores relative to charters:
Here is a summary of these PARCC test score differences between DCPS and charters at those key grades:
While it would be comforting to say that such shifts in the highest PARCC scores are evidence that charters do better than DCPS at the middle school level, and DCPS better at the elementary and high school levels, one cannot make that conclusion because there is no significant change at the state level scores for most of these grades:
This is important to note because of the near parity of absolute numbers of students in charters and DCPS. That is, if one sector was demonstrably having much better test scores overall, you would inevitably see it in the state scores. (We can see, however, that the highest math scores for grade 10 overall are significantly lower than those for grade 8, so one conclusion could be that scoring well on PARCC demands something else in math instruction than is currently given in most DC public high schools.)
Now, if you look at the group of DC public school students with the consistently highest overall PARCC scores–white students–you will see a corresponding demographic shift at those critical grades (4th to 5th, 8th to 10th) that mirrors the same pattern of the test score shift between charters and DCPS for those same grades:
One could also conclude that the generally smaller percentages of higher scores in both charters and DCPS at higher grades (grades 8 and 10) is reflective of a smaller percentage of the highest scoring students taking those tests at those higher grades. Whether that is due to smaller absolute numbers of those students at those grades or attrition over time is unclear, although historically student attrition over time is very strong in DC public education.
Taken together, these results do not suggest that one sector has superior (or inferior) educational strategies or achievements at any juncture–as some have suggested and as some parents clearly believe when they pull their children out of their DCPS neighborhood elementaries to go to charters at 5th grade.
Rather, these results show that demographic shifts in student population affect test scores–shifts that, at least for some Capitol Hill elementaries, have difficult ramifications (and, it would appear, for some charter high schools as well).
Just don’t hold your breath that any education leaders will discuss this publicly–because it gives the lie to commonly held notions of school competition, choice, and test scores as unmitigated barometers of educational virtue.