Forum Index
»
Private & Independent Schools
2011 Admissions Results -- Links to Google doc are in 1st post on page 1
![[Post New]](/jforum/templates/default/images/icon_minipost_new.gif){kind=icon}
Anonymous
| SAM2 If you do this again next year, of please separate SSAT and WSSPI and WISC tests? Many students with 98% IQ's do not score that high on the SSAT! Just a thought ... |
Anonymous
It's not too late for this year even. The survey is just in beta-test form right now, so I can work out kinks like the one you describe. I assumed that SSAT scores come with percentile ranks 1-99, just like other tests, so the same percentile scale would work for all tests. Is that incorrect? Please set me straight. Thanks. Sam2 |
Anonymous
Correct. All are just made-up submissions right now, half of them by me. Please submit more. |
Anonymous
Thanks for your comment. You are absolutely correct that the people posting will be a self-selecting group, and that some people may post fake results, so I don't know that anyone can make predictions based on any of the responses. But I really hope that fake responses don't happen much. This whole admissions thread is about people sharing information, and not some silly contest where schools or parents are competing. There's no way to stop some people from being jerks and spoiling the usefulness of the thread for everyone else, but just maybe they'll go somewhere else for once. If you have any ideas about how to discourage fake responses, I'd be really happy to hear them. |
Anonymous
|
SAM2
Yes, SSAT is percentile. TIA! |
Anonymous
|
There's no way to discourage fake responses. The most you could do, I'd think, would be to attempt to prevent double submissions and, in doing so, you'd both exclude legit posts (e.g. same IP address attributed to two different posters at the same worksite, 2nd one is prevented from posting) and fail to catch anyone who posted multiple times from different IPs (e.g. lots of people I know routinely use 4 different IPs (work computer, home computer, ipad, laptop with integrated wireless) so you don't have to imagine some highly-motivated lunatic running from cafe to library to school to scam your poll. Just a garden variety jerk with a good but not exceptional arsenal of electronics.
And, of course, even if you could somehow limit each poster to one submission, that one submission could be completely false (non-applicants, misrepresented scores or results). I think it really devalues stats to create them in this way. |
Anonymous
| I say do it. We're a sophisticated lot. We know the margin of error is going to be bigger than a methodically sound sampled survey! |
Anonymous
I agree with everything you posted; there's really no way to prevent a determined jerk from contaminating the pool of results. If someone wants to piss in the pool, I can't stop her. But maybe I can emphasize heavily that this survey is about descriptive and not inferential statistics. In other words, it would be completely foolish and false to try to use any of the data to predict how future applications might fare, or to evaluate schools or students in any way. This survey tool is just a tool to tally up the numbers. Maybe that might discourage people from flooding the survey with fake results -- probably not though. |
Anonymous
|
Suggesting that maybe you're less sophisticated than you think. Margin of error increases as sample size decreases, but that assumes a still randomized sample.
In a self-selected sample that may include data that's utterly fictional it makes no sense to talk about margin of error. We're talking pseudo-science at that point. |
Anonymous
| Sorry, should have quoted. 11:49 was a response to 11:38 (not 11:46). |
Anonymous
| 11:49 I was using "margin of error: broadly, it does not only apply to sampling margins of error. The term margin of error can apply to other aspects of survey validity, including bogus responses. |
Anonymous
Thanks for the clarification. But trust me ... I'm likely the least sophisticated one here! Sam2 |
Anonymous
Seriously, what kind of person would do that? It won't change the actual outcome of who gets in and who doesn't. So much has been discussed here about how there is no definitive formula to exactly why a particular kid gets in over another - it's a matter of what is needed to round out a class. My only interest in seeing these results as people get them would be to say - ok, there have been 10 boys accepted for K at Maret - then I could adjust my expectations accordingly while waiting for the mailman. But honestly, at this point aren't we all aware it's out of our hands? What is the motivation of someone to post fake results? We're all anonymous, so it can't be for bragging rights. So, would it just be to inflict intentional pain, stress, harm on those parents still waiting for results? That's pretty sick if you ask me. Aren't we all in a similar boat here? Doesn't everyone want the best for their child and have high hopes for our child getting into the school(s) that they (and we) loved and felt most comfortable with? Why can't we support each other as we wait and support each other if we are disappointed by our results? Parenting is not a contest - I feel really sorry for any children of people that would do something like that because what kind of examples are they showing their kids in everyday life lessons? Ugh.... Can't people just be honest and not jerks?! |
Anonymous
http://en.wikipedia.org/wiki/Margin_of_error http://www.amstat.org/sections/srms/pamphlet.pdf see 63 ff, esp p. 66 Non-sampling errors aren't reflected in margin of error stats. Maybe you meant "metaphorically" rather than "broadly." |
Anonymous
|
Actually I meant broadly, as applied to the results. If I am misapplying the term, my excuses.
margin of error In survey research, an estimate of the difference between a result obtained from a sample (e.g., the sample mean) and the corresponding true population value (e.g., population mean). |