Quote from: Devlyn Marie on March 04, 2018, 01:10:12 PM
It's interesting how the discussion wobbles from "Who is" to "Who should be counted"
I don't think that either of those question sums up the intent of the continued discussion though. Drawing statistical numbers does require a definition to go off of, and so consideration of that definition and how it might be addressed is a natural topic for the question of how many, often for the sake of
inclusion not exclusion. This in turn does flow into the issue of delineation of subgroups to fully explain the data, but then that is just how data works to begin with, as it is meant to be broken down and analyzed in order to be applied for a purpose. Statistical data needs a reason to exist, and a non delineated transgender population value doesn't really have one other than the aforementioned sating curiosity. These and countless other things absolutely
have to be considered in any study or polling in order to create remotely reliable numbers, otherwise you just wind up back where we are now. Self identification is wonderful from a social-cultural perspective, but it severely complicates attempts at representative math.
(Case in point on the importance of actually working these questions out: Someone recently posted a study link in these forums which had a somewhat poorly defined and inconsistent definition of "non-op" presented. The definitions were problematic in a number of ways, both by being inconsistent with one another, between the contradictory synopsis on the study and wording posted elsewhere, as well as not in line with the most commonly accepted definitions.)
(Random addendum because thinking about this is oddly relaxing for me... To go back to the delineation aspect for a moment, think about the overall U.S. population. If you provide the U.S. population, what can you glean from that? Well, pretty much just that number itself. You don't know the ethnic, religious, class, or fiscal breakdown, you don't know who is suffering and who is doing perhaps a little too well. Toss in another stat, such as the popular vote tallies for the Presidency when the candidate who led the popular vote lost the election. Suddenly that lack of delineation in the general population data, really, really, really matters. The overall population number explains precisely nothing of value to anyone. Voting patterns, disenfranchisement, fraud, you name it, all impossible to detect. Add in that delineation, patterns emerge. Answers are given, and the study/poll/census's existence is justified, not least of which from a financial perspective to warrant the commissioning of the study/poll/census in the first place.)
