Active Users:367 Time:17/06/2025 06:58:39 AM
Re: You seem to have perfected whining to a Talent... Tor Send a noteboard - 11/11/2012 08:43:19 PM

So, I went and had a look at some statistics texts, and here are a few juicy quotes (emphasis mine):

---

Self-selection bias is the problem that very often results when survey respondents are allowed to decide entirely for themselves whether or not they want to participate in a survey. To the extent that respondents' propensity for participating in the study is correlated with the substantive topic the researchers are trying to study, there will be self-selection bias in the resulting data.


Nice. But why did you ignore the very next sentence?

In most instances, self-selection will lead to biased data, as the respondents who choose to participate will not well represent the entire target population.


That's exactly what I've been saying. A self-selected sample is not representative of the entire population. 1000 women is too low a sample number when they're self-selected. So you see a bias in one way. The very next 1000 self selected women from the same area might have been biased in the opposite way, or biased even more towards weak women. The point is, in a non-random sample, you'll almost certainly see bias. There is no one reason for this.

---

A voluntary response sample consists of people who choose themselves by responding to a general appeal. Voluntary response samples are biased because people with strong opinions, especially negative opinions, are most likely to respond.

---

This is the case with political poll. In epidemiology (which the current case is closer to, since channeling the OP is not an opinion, it is a genetic and metaphysical predisposition), it is not the strong opinions of respondents that leads to sample skewing.

A sample of convenience is a sample that is not drawn by a well-defined random method. The big problem with samples of convenience is that they may differ systematically in some way from the population. For this reason, samples of convenience should not be used, except in situations where it is not feasible to draw a random sample. When it is necessary to draw a sample of convenience, it is important to think carefully about all the ways in which the sample might differ systematically from the population. If it is reasonable to believe that no important systematic difference exists, then it may be acceptable to treat the sample of convenience as if it were a simple random sample.


Ok, so the point of these quotes was to back up what I've been saying all along, and which you have protested, i.e., that there has to be a specific mechanism which links the probability that a person will volunteer for the test with the characteristic you are testing for there to be a self-selection bias. Personally, I prefer arguments and reason instead of the authority of a book, but since you asked for books, I guess you don't.

But there is reason to know a systematic difference exists. Its not a matter of belief here. We know for a certain fact that this non-random population is showing characteristics markedly different from the norm.


I guess this sums it up. If you are now saying that there is a reason to assume a systematic bias in this particular sample, then I can't argue with that, except to say that in my opinion it seems more likely that RJ made a mistake with one of the two numbers, rather than adding a massive bias to those novices as a plot device.

However, I would be very interested (and I'm not being sarcastic, I would actually be very interested) to hear what you think the mechanism of the bias would be.

Your thought experiment is so stupid I'm at a loss for words. It is not remotely comparable because the probability of a coin landing heads or tails is decided at that very moment, and depends on the property of the coin, not the person tossing it!


But in the other post, you agreed that strength in channeling is essentially a random number assigned at birth, unknown until you are tested. I fail to see how this is different from a bunch of people flipping a coin, marking the result without looking at it, and then at some later point volunteering to have their coin tested.

Edit: Fixed a missing quote-tag
Fram kamerater!
This message last edited by Tor on 11/11/2012 at 08:45:42 PM
Reply to message
The Bell Curve revisited - 29/10/2012 09:44:09 AM 1525 Views
Re: The Bell Curve revisited - 29/10/2012 10:21:27 AM 909 Views
That's incorrect... - 29/10/2012 10:26:49 AM 1482 Views
Re: That's incorrect... - 29/10/2012 10:36:32 AM 912 Views
RJ the physicist didn't know math, so that Shannow could be right... - 29/10/2012 02:11:19 PM 828 Views
Response to a few of your poorly researched points... - 29/10/2012 02:31:17 PM 771 Views
Re: RJ the physicist didn't know math, so that Shannow could be right... - 29/10/2012 02:37:33 PM 789 Views
Exactly... - 29/10/2012 02:39:30 PM 787 Views
there are dozens of reasons for this - 29/10/2012 08:18:18 PM 805 Views
Excellent point. - 29/10/2012 08:24:37 PM 834 Views
Re: there are dozens of reasons for this - 29/10/2012 09:07:35 PM 735 Views
Again I don't argue that genetics play no role - 30/10/2012 01:57:24 AM 712 Views
Re: Again I don't argue that genetics play no role - 30/10/2012 07:07:17 AM 741 Views
I don't think it plays much role in the plot - 30/10/2012 03:17:55 PM 889 Views
Once again just so,we are clear on my stance with Genetics and Strength - 30/10/2012 03:27:11 PM 752 Views
That the 1000 Novices aren't a random sample of the population? - 29/10/2012 08:23:47 PM 689 Views
And why would it be biased towards those with lower strength? - 29/10/2012 09:11:25 PM 695 Views
Absolutely no reason... - 30/10/2012 01:35:35 AM 794 Views
Re: Absolutely no reason... - 30/10/2012 06:43:54 AM 696 Views
Only if it was a random sampling. Which this is not. - 30/10/2012 01:58:34 PM 784 Views
That's exactly the point. I want you to explain why it wasn't random. - 30/10/2012 02:14:59 PM 706 Views
It wasn't random because it was a self-selected sample! - 30/10/2012 02:43:03 PM 721 Views
Re: It wasn't random because it was a self-selected sample! - 30/10/2012 02:47:30 PM 719 Views
Go read a stats text will you? - 30/10/2012 02:54:16 PM 715 Views
Done - 31/10/2012 09:34:11 AM 1451 Views
You seem to have perfected whining to a Talent... - 10/11/2012 10:14:19 PM 976 Views
Re: You seem to have perfected whining to a Talent... - 11/11/2012 11:37:16 AM 811 Views
Re: You seem to have perfected whining to a Talent... - 11/11/2012 07:14:48 PM 675 Views
Re: You seem to have perfected whining to a Talent... - 11/11/2012 08:33:59 PM 1484 Views
Re: You seem to have perfected whining to a Talent... - 11/11/2012 08:43:19 PM 1004 Views
Still nothing? - 10/11/2012 03:33:15 PM 797 Views
Still doesn't explain the difference - 30/10/2012 07:01:53 PM 663 Views
Re: Still doesn't explain the difference - 10/11/2012 10:21:00 PM 741 Views
Yes that totally makes sense - 30/10/2012 08:07:16 AM 833 Views
Thank you! *NM* - 30/10/2012 10:19:15 AM 386 Views
That's not what happened... - 30/10/2012 02:01:52 PM 761 Views
Re: That's not what happened... - 30/10/2012 02:15:57 PM 720 Views
Who said it would? - 30/10/2012 02:44:17 PM 720 Views
let's not mix up "random" and "representative" - 30/10/2012 05:28:09 PM 789 Views
Doesn't mean RJ applied it to his series - 30/10/2012 08:23:29 AM 802 Views
But of course he did.. - 30/10/2012 02:13:07 PM 822 Views
I hate to get into these things - 29/10/2012 05:45:50 PM 864 Views
I would love for you to be right, because it would solve all our problems, but 0 is the challenge... - 29/10/2012 07:56:34 PM 819 Views
In the truest sense, you are probably right that it is skewed - 29/10/2012 08:20:52 PM 845 Views
Overwhelm Lanfear, not match her. *NM* - 29/10/2012 08:26:09 PM 420 Views
Truth is, Moiraine was being overly optimistic... - 29/10/2012 08:39:17 PM 767 Views
You're pathetic... - 30/10/2012 01:20:01 AM 706 Views
The quote isn't specific - 30/10/2012 08:32:36 AM 823 Views
Its highly specific... - 30/10/2012 02:15:38 PM 660 Views
Yet neither of them are at full potential and at least equal a Forsaken - 30/10/2012 03:45:24 PM 1312 Views
Honestly! - 30/10/2012 02:07:37 AM 757 Views
Re: In the truest sense, you are probably right that it is skewed - 29/10/2012 09:10:27 PM 762 Views
Lots of people mean perfectly normal distribution when they say it - 30/10/2012 05:25:35 PM 703 Views
Couldn't the Towers method of obtaining Aes Sedai be to blame? - 30/10/2012 12:04:01 AM 903 Views
Re: Couldn't the Towers method of obtaining Aes Sedai be to blame? - 30/10/2012 09:33:44 AM 828 Views
Are you sure about that? - 30/10/2012 12:03:43 PM 830 Views
Re: Are you sure about that? - 30/10/2012 12:19:34 PM 736 Views
That doesn't seem a coherent narrative to me - 30/10/2012 04:26:25 PM 1054 Views
Sharina did not have the Spark, nor did Nicola - 30/10/2012 05:16:40 PM 845 Views
Re: Sharina did not have the Spark, nor did Nicola - 30/10/2012 05:54:41 PM 725 Views
We do not know if Cadsuane or any of the Forsaken are Sparkers - 30/10/2012 10:33:55 PM 856 Views
you're confusing 2 things - 30/10/2012 04:27:32 AM 971 Views
+1 *NM* - 30/10/2012 09:17:07 AM 832 Views
Re: you're confusing 2 things - 30/10/2012 09:21:39 AM 802 Views
Not true... - 30/10/2012 11:49:57 AM 815 Views
One thing - 30/10/2012 05:23:17 PM 791 Views
That's the problem. The BC RJ has "built" has a minimum and a maximum value - 30/10/2012 05:48:55 PM 809 Views

Reply to Message