danielr wrote:Well if a strong argument could be made, why doesn't someone go ahead and do it?
Probably because people have better things to do with their time...
danielr wrote:The statistics on LSN are, as you say, incomplete and furthermore they are self-reported. LSN profiles often are deliberately not reported accurately, so as to provide for anonymity. Add to that the fact that many profiles are inactive, not updated, or possibly even fictional and the primary data source for arguments about "high-numbered candidates" is completely undermined. LSN is not a reliable source of information, period.
All of these criticisms are legitimate but they don't support the conclusion that we cannot rely on LSN at all. And again, we're just looking at broad trends on LSN here. This isn't a scientific study that will be subjected to peer review.
danielr wrote:Also, even if we grant that LSN has profiles of 20% of all Georgetown admits, we're still talking only 20% of nearly 7,500 applications, and the percentage of high-number applicants on LSN and within the total applicant pool is totally unknown.
It's not totally unknown. See kappycaft1' posts on the relationship between LSN and the total applicant pool. And 20% of 7,500 applicants (1500) is a large enough sample size to draw conclusions here. You're correct there is some self-selection bias here, but if anything that helps our cause because we're interested in top applicants, who are more likely to be on TLS and LSN.
danielr wrote:Furthermore, this cycle is weird as hell. I'm not saying we have to disregard past experience, but we can't forget how whacky things have been across the board at many top schools.
People throw this claim out often but how do you support that? Sure, some people are slightly overperforming, but others are slightly underperforming. Some particular schools are doing things differently this year but in terms of performance this cycle seems pretty similar to past ones, maybe a little better.
OK. 1) Considering that you cut my post into four separate quotes, and that we have all spent a considerable amount of time on here discussing this, I'm going to go ahead and say that it's false that no one has made a strong argument on this issue because "people have better things to do with their time." Your activity on this question (and mine too, admittedly) easily disproves the point.
2) Nobody said our arguments had to have the statistical validity of a peer-reviewed scientific study, but if our arguments are to stand the things we support them with do need some strength. Law School Numbers may indicate broad trends, but that is not enough to support what many have said in this thread, namely that Georgetown is shooting itself in the foot by "lowballing" high-numbered candidates. There is just not enough reliable data there to support that conclusion!
3) Your premise that "top applicants" are more likely to be on TLS and LSN misses the point--you don't know the proportion of high-scoring applicants within that 20%, and you don't know how many top applicants are not
posting their information on TLS and LSN. We have no knowledge of the whole here, only a small sample size with an undetermined relation to the entire sample.
4) If my comment about this cycle being weird is unsupported, then most of the assumptions you have made in your arguments are equally unsupported. I think this cycle is weird, and different; I wasn't intending on providing very much support there--I only wanted to indicate that there is much more uncertainty here than we are so far granting. Applications have not been down in recent years in such high percentages as they are this year. That alone makes this cycle weird, and I think most anecdotal evidence will bear me out here.
With that, I think I've said my peace on this issue. I don't dispute the fact that Georgetown is not as generous with scholarships as other schools, even peers. But I also don't think that just because TLSers (or LSNers) are dissatisfied with the results of the scholarship awards that means top applicants in general are. There are too many fallacies and issues with the data to make any such conclusions.