A check of whether the number of students tested in each grade adds up to the number of students tested in “All Grades” reveals that… it’s so close to perfect. There’s only one obvious inconsistency:
dbn grade year subject n score 22K245 3 2007 ELA 21 661 22K245 4 2007 ELA 1 NA 22K245 5 2007 ELA 24 655 22K245 All Grades 2007 ELA 1 NA
(The issue is that 21 + 1 + 24 is not 1.) Something weird is going on here; the number for fourth grade could be wrong too. Who knows? The moral of the story is: don’t use the provided for “All Grades” – add it up yourself.
One interesting thing is that very occasionally a school will test some students in one subject but test no students in the other.
dbn year Math ELA 09X413 2006 NA 31 75K077 2006 NA 2 75X012 2010 1 NA 75X012 2011 1 NA 84K701 2006 102 NA 84M704 2006 164 NA 84Q704 2006 293 NA
It’s a little interesting that three of these are from two District 75 schools, which do special education exclusively. Many of their students don’t take these tests, as I understand it. Three more of these records are from Charter schools, which is indicated by “84” in their district space. Perhaps they just didn’t have their act together in 2006, or had some exemption from the ELA exam? I don’t have any particular theory for 09X413. There are probably many more cases such as this at the school grade year level; I’m not too concerned about it for the moment.
At the school year level, the number of students tested in ELA and Math is exactly the same for only 14% of records. To get a feel for how the number tested in Math relates to the number tested in ELA, we can plot these numbers against each other, as also shown at top:
Where there are differences it’s usually more students taking the Math test than the ELA test. One possible reason for this could be English Language Learners who are somehow “excused from” the ELA test but still take the Math test. Perhaps this was allowed particularly in 2006, which seems to have been a free-for-all when it came to not testing students in ELA. One wonders if schools ever make any sort of effort to keep certain students from testing. At least the numbers are close, especially for the last several years. The persistent outlier even in 2011-2013, clearly apparent toward the bottom of those graphs, is 30Q235:
dbn year Math ELA 30Q235 2006 222 3 30Q235 2007 251 1 30Q235 2008 156 1 30Q235 2009 174 3 30Q235 2010 239 58 30Q235 2011 260 81 30Q235 2012 225 86 30Q235 2013 213 72