<p>For those of you unfamiliar with my school's story (Lexington), nine people applied EA to MIT. Two got their acceptance letters on Saturday at approximately 11:00 am. The third acceptee just got her letter today. No defer/denies yet, however.</p>
<p>Seems the mailing is indeed random.</p>
<p>Random? Is that sarcastic?</p>
<p>We <em>just</em> learned a new technique in stats that allows me to say that the chances of 3 people recieving acceptance letters while none recieving any other sort, given that the chances of acceptance were equal for all people, is .0674. So about 6%. Not very random to me, I'd say the accepts arrived first.</p>
<p>That is because the tubes might have been sent PRIORITY mail.</p>
<p>I just did a postage calculation for my tube. 3.85. According to the USPS website, Parcel post is 3.75 while priority is 3.85.</p>
<p>Priority promises 3 day delivery, under most conditions.
First class mail would take longer.</p>
<p>Which is why very few Defer/Deny envelopes have arrived.</p>
<p>Well there goes my hope.</p>
<p>Yeah, but at least we have some kind of logical explanation. That's pretty interesting.</p>
<p>mmhm. Spartan's post doesn't say anything about deferrals/denials not being PRIORITY mail. I believe benjones has also noted that both letters were sent with the same priority. I think we are over-analyzing.</p>
<p>Gratz Spartan!</p>
<p>This is not a true statistical analysis, as your sample is completely biased and nonrandom. Therefore, any deductions and inferences are NOT reliable. </p>
<p>We don't have the data for the rejects. As I've said before, rejected/deferred applicants aren't bothering to join CC and post their rejection. The result is a very skewed sample which tends towards the many students who have recieved acceptances and joined CC. The rejected and deferred students haven't bothered to join CC, so we cannot justifiably infer that the tubes arrive faster.</p>
<p>mognoose - insufficient data and not regularly occuring data. inconclusive.</p>
<p>spartan - my friend confirmed the priority. However, parcel post should cancel out priority due to hand sorting. And in any case at all, I should have received something by now.</p>
<p>Still, mognoose's 6% ultimatum is disheartening.</p>
<p>hahahhaha, go figure MIT applicants would analyze results statistically and then maul themselves with facts and figures.</p>
<p>But can we help it?</p>
<p>lol really it's no use analyzing anything when you have limited information. it's bound to have a HUGE percent error really and besides the assumption that everyone has the same chance of acceptance does not hold true. How can an RSIer have the same chance of getting accepted than a normal kid with regular stats? Also it is hard to define the chance of getting accepted because there is a human factor that is inmeasurable thus the premise of your calculation is flawed and thus the result is flawed as well.</p>
<p>Crap I totally messed that up anyways. I was thinking of a totally different thing; but thanks for nitpicking the heck out of it anyhow ;)</p>
<p>I was thinking that would be the chance that no more/no less than 3 of 9 would "win" the .12 odds of getting in. The chance that the first 3 observations of 9 all get in is just .12^3, which is an even more disheartening .00178, or about two percent. EDIT: make that two <em>tenths</em> of a percent. uh oh.</p>
<p>What I'm saying is that because this number is extremely low, you'd imagine the odds would balance out to a more reasonable 1 or 2 in 9 getting in, or at least move in that direction (law of large numbers for any statistics people, even though 9 is hardly a large number, but you get the point). But because this number is so low, I would place my bets that fewer of the last 6 would gain acceptance if the first 3 of 9 have.</p>
<p>Also, this is a rough model saying that all applicants have .12 of getting in, which we know isn't true, but for the sake of simplicity. If we knew the SATs, gender, race and "hook factor" of these kids then maybe we could do something better, but come on... work with me people :)</p>
<p>mognoose:
I stand by my initial point (which danielsuo seems to agree with)--your inference is not justifiable for the reasons previously stated.</p>
<p>In addition to my previous points, the CC group is NOT representative of the whole population of applicants. Therefore, the 12.2% rate won't even apply to this scenario. </p>
<p>Spartan Pho3nix:
I really hope your deduction isn't true, despite it making complete intuitive sense. Did you calculate based on the number of accepts and the total MIT spend (they gave a figure for the cost of postage), or does the letter itself have the price spent? Also, I doubt Ben would lie to us about the tubes being sent a different priority. I mean, there are at least 3-5 people on the roster whom we'd definitely expect to see getting acceptance letters. I think we still have hope :-S.</p>
<p>I think it's ridiculous to try to make this kind of analysis.
Ben said that everything was mailed with the SAME priority.
What would make sense to me is that there are still a lot of tubes out there that are taking longer because they have to be sorted by hand.
At a certain point, maybe not having anything yet becomes a sign in your favor - envelopes are sorted by a machine and should be delivered already; I wouldn't be surprised if people aren't posting when they're deferred or rejected.</p>
<p>kimbrel:
I couldn't agree with you more. What else do we need--proof they were sent the same priority? </p>
<p>Also, I was wondering what you mean by the tubes being sorted by hand and the envelopes by machine. This doesn't seem to make sense. You seem to be familliar with the sorting and routing process, please elaborate. I thought the letters were routed by hand (didn't make sense, considering the millions that go out). Surely, they don't have some sort of OCR wizzard where they can actually recognize handwriting on envelopes and sort them accordingly. I'd assume they have the same system in place for handwritten/addressed letters as they do for print-labeled ones. Anyway, you seem to be experienced. Please do explain.</p>
<p>From what i have heard, as long as the address on the envelope is typed, the machines can recognize and sort them. If the address is hand written, it goes into another pile and is hand sorted. The tubes probably won't fit in the machines, so I think they get hand sorted too.</p>
<p>Oh...Interesting.</p>
<p>So, they are using some sort of OCR wizard. In this case, deferrals and rejections should have arrived earliest. This surely can't be the case. Any explanations?</p>
<p>Ben didn't want to defer us, so he was still tenaciously fighting until Tuesday. So they were mailed late?</p>
<p>Hopefully ;-).</p>
<p>The post office actually has handwriting recognition machines. That's why you have to pay extra money if the address is written in somewhere other than the norm. All postcards/letters are handled and sorted by machine.</p>
<p>Are you serious! That's so cool! I wonder where I could get a copy of that software--that way I could convert my notes to a digital format. I thought OCR software was relatively primative--considering it isn't even perfect for printed text. I guess I'm using the wrong software.</p>