A little delay since the last two posts (Intro & Chapter 1, Chapter 2)… Christmas vacation = no preschool, so life has been very full of other things. Also, took a little trip. Here’s my view as I write this morning:
In Chapter 2, the authors introduced their five “essential supports”: 1) school leadership, 2) parent-school-community ties, 3) professional capacity, 4) student-centered learning climate, and 5) instructional guidance system. Note that these are supports, as in they provide conditions that “substantially influence” (p.79) the work of the school, but they do not directly cause the improvement. This is the nuance of a systems approach. The authors also make the point that they are essential and will use a quantitative methodology to show that improvement stagnates without them.
Because of this systems approach, the idea of “holding other factors constant” doesn’t make sense. If each support is reinforcing (or undermining) to the other, holding others constant doesn’t actually give a sense of how the two systems interact, sort of like trying to understand how a steering wheel functions independent of the wheels. This also means that statistical approaches that are designed to control for particular variables don’t work. Thus the approach that is used is “a form of analytic spiral” (p.80). Basically the authors use a large longitudinal database of surveys and test scores to explore these supports. I would be interested to know all the other ideas they tried before coming up with their final analysis. It comes across quite straightforward, but the process was no doubt complex.
Again, they use reading and math test scores, but they use them only as an indicator of improvement if they were in the top quartile or stagnation if they were in the bottom quartile. This approach makes sense to me as the top and bottoms are obviously showing improvement or stagnation whereas those in the middle are harder to parse. Perhaps as my quantitative fluency improves I will have a more critical eye to their methods, but for now I will take it as presented.
A strength of their analysis is that they present both the schools that are improving and stagnating. This bolsters their argument because it shows that schools with high levels of the supports are more likely than chance to show significant improvement whereas schools with low levels of the supports are more likely than chance to show stagnation.
Most interesting to me was the cumulative effects of the supports. Through an aggregated indicator score for the supports compared to improvement in math, reading, and attendance, the authors show a distinct correlation with strength or weakness in the supports.