
To keep anonymity, how about providing percentile score to the nearest integer, so for example, report 99%ile, even if score is 99.4%ile or 99.9%ile. That should be harder for schools to track. |
I definitely support that approach. Right now, I set up options to report percentile in large 10-point increments (e.g., 80th to 89th percentile), with only one smaller increment at the top for 99%. What you're proposing is making an even finer distinction that I currently plan to use. I welcome input from others about the best increment. The people who really should be providing input are those here who are actually waiting on early-March decisions. Those are the ones who hopefully will be responding, so they should say that level of specificity makes them comfortable. Sam2 |
Yes, you are misapplying the term in two ways. Margin of error is a stat that indicates how closely the results of a randomized sample are likely to approximate results from the entire population. This means that margin of error (1) doesn't include non-sampling errors and (2) can't be computed unless you have a randomized sample. |
So no survey but a randomized one has a calculatable margin of error?
|
Right -- and when you can compute it, it only reflects errors associated with sampling (e.g. It doesn't take into account the possibility that someone will lie or miscode data). |
Any last suggestions for ways to improve this counting tool? I plan to finalize it this weekend, and post a "live" version some time next week.
Beta version of survey: https://spreadsheets.google.com/viewform?form...NaSlMtdlRPd3hORmc6MA Beta version of results: https://spreadsheets.google.com/viewanalytics...NaSlMtdlRPd3hORmc6MA Possible adjustments: Any schools to add? Any schools to eliminate? Other student/family characteristics to be tracked? Any to eliminate? Adjust instructions some questions, to remove any ambiguity? Add a question about handwritten notes from Maret, since that seems to be a hot topic every year? Others? TIA. |
Thanks, the links are not working.
|
Don't know about others, but I can't get in to either doc through the links. |
Most of the Virginia schools are missing. I would add: Alexandria Country Day School Browne Academy Burgundy Farm Country Day School The Langley School |
Oops. I must've mis-copied. Try these: Beta survey: https://spreadsheets.google.com/viewform?form...NaSlMtdlRPd3hORmc6MA Beta results: https://spreadsheets.google.com/viewanalytics...NaSlMtdlRPd3hORmc6MA |
Will add. Do you think there are appreciable numbers of people on DCUM applying to those schools? I'm not too familiar with them, so maybe I just don't notice people posting about them. |
Great survey, SAM2!
I would add "First Choice Letter?"--since this seems to come up on a regular basis. |
We applied to one of these schools... |
I agree on first choice letters and the handwritten waitlist point.
TIA! |
SAM, YOU ROCK! This is awesome!! Signed, Big Fan |