cyberdad wrote:
uncommondenominator wrote:
There are some cases when small samples are viable - usually when the population being studied is equally small. But when the population is literally the general population as a whole, a much larger sample is needed. 443 people is like, one ten-thousandth of one percent of the population of America. Please, explain to me, in detail, how that sliver of people, must surely be an accurate cross section of all 330 million people. Or all 25 million Australians. Or all 67 million Britts. Or the 8 billion people of the world.
Ummm your reasoning really doesn't make sense? you earlier commented that in order for studies to be taken seriously you need a sample of thousands. But then you flip and say sometimes small samples are viable, But then you backflip again and say a sample of 443 is miniscule and not representative of the wider population??
Please go back and open your dusty unread statistical textbooks and re-familiarise yourself with central limit theorem and statistical power. Read about experimental design. You don't need a sample larger than even 30 if you are using techniques like stratified random sampling,
But anway this isn't about stats, it's about a desire to undermine
the well established premise that humans are driven by a biological urge to seek companionship/friendship with people much like themselves which include "body size".
Me sorry. Me talk more slow. Make easer, for you read.
If you actually paid attention, you'd notice that I said that studies with a LARGE population need a LARGE sample, and that studies with a SMALL population can get by with a SMALL sample. No backflipping required. A sample of 30 may be meaningful as related to a population of 500 - but not so much to a population of 500,000. Or 50,000,000. I'm not sure why that's such a difficult concept. But kudos to you for finding a way to make it sound silly. I bet you feel very "clever".
And thank you for finally getting around to bringing up n>30. How long did it take you to google something to use as a supposed "exception" to fling and flee. The idea of n>30 even applies to coin tosses. We learn that in the first weeks of stats.
Ooh, stratified random sampling, eh? What website did you regurgitate that from? Again, if you READ, stratified random sampling involves dividing people into SMALLER SUBGROUPS. As mentioned above, by me, smaller populations can use smaller samples. The key element you seem to be missing is, conclusions drawn about subgroups, only apply to those subgroups, not the greater population. It's good for studies like "low income people in small towns are more likely to have unhealthy eating habits compared to high income people in large cities", but not so good for claims like "everyone thinks fat people are uggo".
In some cases, such as researching individuals with superior autobiographical recall, they are rare enough that there aren't even 30 people in the population, of which to select a sample. There is no choice but to select fewer than 30, since there aren't even 30 to select from.
You can make "clever" comments about how "dusty" my "unread" stats book must be, but it's just talk. Yakkity yak.
The 2x2 ANOVA studies I designed for my research professor seemed to meet with his approval. Graduate-level Cognitive Behavioral Psychology and Radical Behaviorism gave me quite a bit of insight as to how psych and behavioral studies are designed, performed, and progressed. I think I'll trust the PhDs who have accredited me over the random internet person who "read stuff" and "noticed things", and who has every reason to paint me as ignorant, for their own benefit.
Actually, this is about dad bods, and whether women like them.
Even if we take your "like dates like" stuff to the absolute extreme, that still doesn't mean that there aren't big ladies, who prefer big men, because "like dates like". So the answer to "do some women like dad bods" is still YES - even if those women are themselves a little yummy around the tummy. Fat people happily dating other fat people IS "like dating like" - your personal issues with the existence of obesity are an entirely separate issue, and other people's life choices are really none of your business.
Furthermore, as long as you're still here, I really would like an explanation of your comment about "privilege" and being NT...
Last edited by uncommondenominator on 06 Mar 2023, 8:30 pm, edited 2 times in total.