Re:Diversity of the Mind
by dragonwriter on Monday January 05, 2015 @07:30PM
(#8842238)
> Does it matter precisely what "liberalism" is when 98% of the current grad students identify themselves as such?
Nothing in the source indicates that. There is a claim -- which cites a personal communication with no identified methodological basis -- that 2% identify as conservative (given a set of options that, from the parallel listing of the numbers from a broader population, appears to be liberal/moderate/conservative, and so which does not, even if accepted as ironclad, support a 98% liberal conclusion -- and further, the subsample size of "graduates and postdocs", and therefore the error margins of that group, are not identified.) But, in any case, yes, it matters what it means (and, particularly, whether it means the same thing to all the people with that self-identification) if you are interested in whether there is diversity of ideology rather than diversity of ideological identity labels. It is a fairly massive error to assume that those mean the same thing.
> Do you really think you're going to be able to slice and dice words until this isn't a problem anymore, in the face of such a percentage?
If it doesn't mean even roughly the same thing to the various people, what a large advantage in self-identification may tell you is that "liberal" is a popular label, without telling you that the label has any shared meaning which is itself overwhelmingly popular.
> It defies belief that this is just the error bars talking.
That would only be true if we had the size of the relevant subsamples and thus could have any idea of the size of the error bars, all we have is the full sample of 292, but that includes more than just the subgroup with 2% conservative identification for which you have invalidly assumed a 98% liberal identification.
> And especially so for social scientists... are they really the people you want to accuse of being unclear on the meanings of the words "liberal" and "conservative" in some context?
As someone with a degree in political science, I am quite aware that those terms have lots of different meanings even in the context of political ideology, which is why -- when actual substantive ideology rather than ideological identity is studied (and the "liberal" and "conservative" labels are used in the study), the actual operationalizations used of "liberal" and "conservative" are radically different between different studies.
Nothing in the source indicates that. There is a claim -- which cites a personal communication with no identified methodological basis -- that 2% identify as conservative (given a set of options that, from the parallel listing of the numbers from a broader population, appears to be liberal/moderate/conservative, and so which does not, even if accepted as ironclad, support a 98% liberal conclusion -- and further, the subsample size of "graduates and postdocs", and therefore the error margins of that group, are not identified.) But, in any case, yes, it matters what it means (and, particularly, whether it means the same thing to all the people with that self-identification) if you are interested in whether there is diversity of ideology rather than diversity of ideological identity labels. It is a fairly massive error to assume that those mean the same thing.
> Do you really think you're going to be able to slice and dice words until this isn't a problem anymore, in the face of such a percentage?
If it doesn't mean even roughly the same thing to the various people, what a large advantage in self-identification may tell you is that "liberal" is a popular label, without telling you that the label has any shared meaning which is itself overwhelmingly popular.
> It defies belief that this is just the error bars talking.
That would only be true if we had the size of the relevant subsamples and thus could have any idea of the size of the error bars, all we have is the full sample of 292, but that includes more than just the subgroup with 2% conservative identification for which you have invalidly assumed a 98% liberal identification.
> And especially so for social scientists... are they really the people you want to accuse of being unclear on the meanings of the words "liberal" and "conservative" in some context?
As someone with a degree in political science, I am quite aware that those terms have lots of different meanings even in the context of political ideology, which is why -- when actual substantive ideology rather than ideological identity is studied (and the "liberal" and "conservative" labels are used in the study), the actual operationalizations used of "liberal" and "conservative" are radically different between different studies.