cocoj
I wonder if this same sort of analysis on GMC data could be run for schools that DO publish admissions statistics. Then we could compare INSEAD's GMC interview-to-admit conversion to other schools and adjust for the self-selection error.
Anyone bored?
IIRC there aren't many schools that give breakdowns between quantity of *all* of the following:
1. Applicants
2. Interviews granted
3. Places offered
4. Matriculations
1 and 4 yes, but not the middle two which is effectively what I was looking at -- and applies directly to those still interested in this thread

. I think it's possible to back-calculate #3 from #4 using assumed yield %s (which folks attempt to publish but are based on weak foundations to begin with). However, #2 is the tough one in the bunch.
In this case, we have what I would categorize as consistent figures (one source, one time period, one "subject" - INSEAD) and anecdotal evidence from the horse's mouth establishing a lower bound to the interview-to-admit rate ("above 50%" --
https://gmatclub.com/forum/insead-january-2011-intake-89459-420.html#p724606). Unfortunately GMC doesn't have large enough sample sizes pre-Jan11 intake to make it a multi-time-period analysis. Instead of synthesizing a more accurate range by applying a subjective "self-selection" adjustment, it's probably best to interpret the horse's mouth statement...Additionally, one school's GMC-to-published stats variance would not necessarily be analogous to any other school's, so you're comparing apples to oranges because you're not sure your apple is really representative of all other apples out there. By all means I encourage someone to try to stitch the analysis together, but I'm sure it will end up, pardon the pun, fruitless...or worst case lead to misleading conclusions.
IMO, with its "above 50%" statement (read Jan 2010 Intake thread and elsewhere), INSEAD reinforces the exclusivity of the interviewee pool without devaluing it at the same time -- which it would do if it actually said "around 70%". INSEAD could easily say "around 60%" or "typically between 40-60%" if that were actually the case, but they leave the upper bound up to interpretation (indicating that it is in fact well above 50%, contrary to the school's perception and marketing strategy as an elite, selective/competitive institution). Typically you provide a range if its bounds are reasonable and don't devalue your position, but in this case they give a benchmark figure with an unlimited upper bound (well, ok, 100% is max) in the hope that they will effectively distract most with the 50% and quash follow up questions.
They're saying, "you guys are already pretty exclusive since we eliminate quite a large % of applicants in the first round, but it's still a very competitive group at the interview stage, so stay on your toes, you're not there yet" -- they *don't* say "our applicant pool isn't as competitive as we make it out to be, so once you make it to the interview, you've gotta be brain dead to not make it". Not that it's as simple as that, but there is a not insignificant psychology / strategy behind selectively providing figures to a very numbers-hungry population and retaining one's marketability at the same time.
I am convinced INSEAD's insistence on not reporting any admissions statistics implies a less-than-competitive (or at least, in its eyes, *qualitatively* comparable to other top-tier institutions) process than it markets itself as having. Not that INSEAD is not competitive or top-tier, but it has certain prereqs in place that effectively discourage higher volumes of applicants that would otherwise typically enhance its selectivity (see: language & international requirements that other schools don't have). What we have here is a very self-selecting institution (read: high acceptance rates) marketing itself as internally selective to boost / maintain reputation amongst those who look exclusively at numbers (FT, BW, Forbes, Economist, and the rest of the rankings folks).
Just the ravings of a conspiracy theorist I guess...
I hope we get some more interview posts soon so we can compare S'11 intake to J'11 to prove/disprove my point. At this stage we'll need quite a few more to get up to the +/- 40 range.