Top Schools Rankings: The Math
Tuesday, May 11, 2010
How'd we do it?
Carefully. We gathered hundreds of pieces data related to school quality: reading, math, and writing test scores (both the New England Common Assessment Program, or NECAP, scores, as well as SATs), student-teacher ratios, spending per pupil, and graduation rates for 52 public, charter, and technical schools in the state.
What followed was precise statistical analysis, guided by a methodology used in similar rankings created elsewhere in New England. After collecting the relevant data, we calculated the average values in each of the categories and the degree to which each school either exceeded or failed to reach those averages.
Those deviations from the average were standardized so that different categories could be compared meaningfully, and then we used a weighting formula to give certain categories more importance than others. We wanted, for example, a school’s student-teacher ratio to matter more in our ranking than its Math SAT scores – though test scores all together account for 60% of the weighting.
Each school’s weighted numbers were added into a single evaluative number, which, when ordered from highest to lowest, gave us our ranking.
Enjoy this post? Share it with others.
Commenting is not available in this channel entry.
Comments:
ken fish
11:21am on Tuesday, May 11, 2010
is there more precise info available on your 'weighting' and 'standardizing'?
the selection of criteria and their relative weighting determine the outcome of the rankings....and reveal the values of the evaluators. i'm not yet sure that this isn't another useless exercise in 'ranking schools'!
how about using student GROWTH, rather than absolute performance? this is closer to identifying the contribution of the school to a child's success.
Damian Ewens
3:40pm on Tuesday, May 11, 2010
The Math?? The grammatical mistake in the first sentence doesn't convince the reader that your study was done carefully. This explanatory paragraph allows one to not understand how you did it but simply to get a vague idea. If it is precise and guided by methodology then show us.
But this isn't really about the math. Nor wether these schools are actually good or bad. This appears to be another excitable list to get readers to look at blinking Dunkin Donuts ads. If you are trying to compare schools, for instance, you might want at least one column for the types of students or teachers that are in the schools. Maybe you did. The reader can't tell. This is simply another futile exercise portraying our acceptance of standardized tests as accurate measures of school success.
Take a look at Dr. William Sedlacek's Non-Cognitive Variables as a strong indicator of student success in life if you are interested in finding out about a path forward, and away, from tests and other poor data points.
Damian Ewens
http://williamsedlacek.info/publications/surveys/noncognitiveadmissions.html
Andrea Jospeh
3:24pm on Wednesday, May 12, 2010
Mr. Ewens,
If you are writing to criticize others' grammar, then your own grammar and spelling should be correct. Last time I checked, a list is an inanimate object, which is not excitable. The list might excite others to a response, which is I believe what you intended to say. I would also argue that the list does not "portray" but rather promotes acceptance of standardized tests as accurate measures.
Damian Ewens
4:20pm on Wednesday, May 12, 2010
Perhaps Andrea, though I'd suggest that there is a significant difference between a quick blog comment and a published article claiming 'precise statistical analysis' on a complex topic.