One other interesting (?) point: Table Seven in the Report is titled "Connecticut Academic Performance Test - Five Year Trends (%)." The Table is on page 24. The interesting point is that it only covers four school years!
______________________________________________________________________________
NARROWING GAPS (3-8): AN ANALYSIS OF FIVE YEARS TRENDS
Subgroups
1. Point six should be the first point, as it establishes the issue to be examined. The issue is not whether performance gaps exist among the subgroups. That is obvious. The issue is whether past actions are actually closing the gaps. From the data presented, it is impossible to answer this question because there is no definition of what a gap is, no discussion of the size of a gap, and no analysis showing the rate of closure or widening of a gap.
2. I offer what could be used as a measurement tool. Since the state ascribes some meaning to the Goal level (and I have already discarded Proficient as a meaningful target level), I will use Goal as the basis for analysis. Does it represent the level which will guarantee success in college or on the job? That is a larger debate than I care to undertake here. With that side-stepped, what should that goal (small “g”) be? Nothing less than 100% of all subgroups at Goal level. Achievable? Probably not, but it is a useful construct against which to measure progress.
3. Based on the data, we need to determine if the trend in Goal percentages is up, down, or flat (is the slope of the trend line positive or negative) and how fast is progress being made for one group versus another (what are the relative slopes of the trend lines)? We also need to determine if the fit of the line is at least reasonable (correlation coefficient).
4. Applying this to the data in Table 2 yields:
| Goal Percentages | ||||||||
| 2006-07 | 2007-08 | 2008-09 | 2009-10 | 2010-11 | SLOPE | Intercept | Corr. Coeff. | |
| YEAR | 0 | 1 | 2 | 3 | 4 | |||
| District | 79.7 | 78.3 | 81.1 | 81.6 | 81.8 | 0.75 | 79.0 | 0.80 |
| Asian | 91.8 | 87.9 | 92.1 | 91.9 | 94.0 | 0.84 | 89.9 | 0.60 |
| Black | 42.4 | 32.3 | 43.0 | 45.1 | 44.8 | 1.76 | 38.0 | 0.53 |
| Hispanic | 56.8 | 54.2 | 60.0 | 59.0 | 62.3 | 1.58 | 55.3 | 0.81 |
| White | 83.4 | 82.8 | 84.7 | 85.8 | 85.9 | 0.80 | 82.9 | 0.91 |
| Female | 79.8 | 77.8 | 80.0 | 80.9 | 81.6 | 0.67 | 78.7 | 0.74 |
| Male | 79.6 | 78.7 | 82.1 | 82.2 | 82.0 | 0.83 | 79.3 | 0.80 |
| SPED | 39.1 | 38.9 | 44.4 | 44.5 | 42.8 | 1.30 | 39.3 | 0.74 |
| ELL | 54.7 | 53.5 | 65.9 | 59.5 | 60.9 | 1.84 | 55.2 | 0.58 |
| FRD | 47.5 | 40.1 | 48.4 | 50.3 | 49.3 | 1.38 | 44.4 | 0.54 |
5. First the good news. The correlation coefficients are reasonable, so we may get some useful information out of this. The slopes (which can be viewed as the annual gain in percentage points) are all positive, so all of the subgroups are trending up, which is great. The slopes of the Black, Hispanic, SPED, ELL and FRD subgroups are all much higher than the district, White or Asian slopes, indicating that (at least by this measure) the relative gaps are closing.
6. Now the bad news. If the goal is to get 100% of each subgroup to Goal, it is going to take a very long time based on the progress being made, or more pointedly, based on what we have been doing. ELL will get there in 21 years, and FRD in about 37 years.
7. While the methodology may be open to criticism, the real take away is that if we truly want to make a difference, we need to make a radical change, not the incremental change we have had.
8. For point 7, the one point which quotes data, the conclusions (“there is no gap in performance of males and females on CMT mathematics at all levels” and “what is important…is that the overall gains are evenly matched over the last five years”) are debatable. To conclude that a 4.5% spread is not a gap is to ignore the actual number of students this percentage indicates. At roughly 625 students per grade across six grades, and assuming a 50/50 female/male split, means that about 84 more females than males are below Advanced. That is almost four full classes. And contrary to the second conclusion, the gap is widening by five females a year (because the actual percentages are 3.4% growth for males and 2.2% growth for females).
Benchmark Districts
1. As was shown above, the elementary grades are not closing any gaps between Greenwich and DRG A and B districts. Even a cursory look at Chart 14 (page 20) shows that the only gap being closed is at the Goal level with DRG A, and a simple calculation shows that the gap will be closed three years after DRG A reaches 100%, that is, DRG A is so far ahead (almost 10%) that it is difficult for them to gain the same amount each year as Greenwich. And it should be noted that Greenwich gave up the 1.5% lead it had over DRG B at the advanced level.
No comments:
Post a Comment