Posts: 10,465
Threads: 197
Joined: May 2016
(05-20-2016, 09:25 PM)Kinser79 Wrote: (05-20-2016, 04:23 PM)Odin Wrote: (05-20-2016, 08:53 AM)Kinser79 Wrote: A sample size of 196 is not indicative of anything. The size of the sample is far too small. They may have only found a couple dozen extremely stupid people. This needs to be replicated with a much larger sample size.
Are you implying it won't? Often a smaller study is done to see if a full study is worth doing.
Not at all. There seems to be no limit to what universities will waste money on these days. I'm just saying that it doesn't mean a thing unless they have a relatively diverse sample size that is at least 10 times larger. One has to account for drawing a couple dozen incredibly stupid people in survey style 'experiments' of this nature. Surely you can see that having just 10 such persons can swing the findings in a wrong direction far more than a standard deviation.
Professors operate under the dictum "Publish or perish". They have good cause to do research, and just about anything that has some plausibility will be studied.
So you don't like the consequences of successful research? So long as the research doesn't have amoral methods it is OK.
Universities aren't solely places of teaching. Research is part of the teaching, at the least for graduate students.
The ideal subject of totalitarian rule is not the convinced Nazi or the dedicated Communist but instead the people for whom the distinction between fact and fiction, true and false, no longer exists -- Hannah Arendt.
Posts: 1,499
Threads: 16
Joined: May 2016
The morality of the method isn't the problem, I don't think a voluntary survey could have such problems, rather it is the smallness of the sample size or have you not been reading the posts in this thread? Or do you simply nut understand that when one conducts surveys that sample size is of paramount importance.
With a sample size of 200 one person sways the figures 0.5%. 20 people sway the figure a full 10%. Standard deviations usually rest in the 2.5-5% range, so merely finding 12 people in this study who are incredibly stupid skews the numbers by one full standard division.
Of course we are talking about statistics here which requires a minimum knowledge of basic maths, so....
Posts: 1,216
Threads: 29
Joined: May 2016
(05-20-2016, 10:04 PM)pbrower2a Wrote: So you don't like the consequences of successful research? So long as the research doesn't have amoral methods it is OK.
He doesn't like the results so he's trying to rationalize why the results don't matter. 
Posts: 1,216
Threads: 29
Joined: May 2016
05-21-2016, 11:43 AM
(This post was last modified: 05-21-2016, 11:45 AM by Odin.)
(05-20-2016, 09:25 PM)Kinser79 Wrote: Not at all. There seems to be no limit to what universities will waste money on these days. I'm just saying that it doesn't mean a thing unless they have a relatively diverse sample size that is at least 10 times larger. One has to account for drawing a couple dozen incredibly stupid people in survey style 'experiments' of this nature. Surely you can see that having just 10 such persons can swing the findings in a wrong direction far more than a standard deviation.
You think preliminary studies are a waste of money? REALLY??? The opposite is true, they are done because the folks giving out research grants want to see if paying for a full study is worthwhile and not just a waste of money. These smaller studies are done exactly because money is tight, not because there is a lot of money to waste.
Posts: 1,499
Threads: 16
Joined: May 2016
(05-21-2016, 11:37 AM)Odin Wrote: (05-20-2016, 10:04 PM)pbrower2a Wrote: So you don't like the consequences of successful research? So long as the research doesn't have amoral methods it is OK.
He doesn't like the results so he's trying to rationalize why the results don't matter.
(05-21-2016, 11:43 AM)Odin Wrote: (05-20-2016, 09:25 PM)Kinser79 Wrote: Not at all. There seems to be no limit to what universities will waste money on these days. I'm just saying that it doesn't mean a thing unless they have a relatively diverse sample size that is at least 10 times larger. One has to account for drawing a couple dozen incredibly stupid people in survey style 'experiments' of this nature. Surely you can see that having just 10 such persons can swing the findings in a wrong direction far more than a standard deviation.
You think preliminary studies are a waste of money? REALLY??? The opposite is true, they are done because the folks giving out research grants want to see if paying for a full study is worthwhile and not just a waste of money. These smaller studies are done exactly because money is tight, not because there is a lot of money to waste.
Seems someone doesn't understand English as well as statistics.
The fact of the matter is that if they wanted a preliminary study then they should have actually conducted a preliminary study that could give them remotely accurate results--which would require a sample size that ten retards couldn't throw off by a standard deviation. Which is how actual science is done. But since this is a social science which is neither social nor science I don't really think it matters.
As for wasting money, conducting a smaller study that gives inaccurate results is precisely wasting money.
Posts: 149
Threads: 1
Joined: May 2016
05-21-2016, 06:32 PM
(This post was last modified: 05-21-2016, 06:33 PM by TnT.)
(05-21-2016, 12:19 PM)Kinser79 Wrote: Seems someone doesn't understand English as well as statistics.
The fact of the matter is that if they wanted a preliminary study then they should have actually conducted a preliminary study that could give them remotely accurate results--which would require a sample size that ten retards couldn't throw off by a standard deviation. Which is how actual science is done. But since this is a social science which is neither social nor science I don't really think it matters.
As for wasting money, conducting a smaller study that gives inaccurate results is precisely wasting money.
Actually, if one truly understands statistics, it turns out that obtaining a good sample is probably the most difficult task. If the sample truly represents the data points being studied, then 196 could very well be enough for useful information. Clearly there are a lot of thngs to worry about in a statistical study - the SIZE of the sample is one of those, but certainly not the most worrisome.
Besides, especially given the fact that in our society we have lots of Republicans, it might well be a legitimate finding to have "ten retards" or more in the sample.
[font=Arial Black]... a man of notoriously vicious and intemperate disposition.[/font]
Posts: 1,499
Threads: 16
Joined: May 2016
(05-21-2016, 06:32 PM)TnT Wrote: (05-21-2016, 12:19 PM)Kinser79 Wrote: Seems someone doesn't understand English as well as statistics.
The fact of the matter is that if they wanted a preliminary study then they should have actually conducted a preliminary study that could give them remotely accurate results--which would require a sample size that ten retards couldn't throw off by a standard deviation. Which is how actual science is done. But since this is a social science which is neither social nor science I don't really think it matters.
As for wasting money, conducting a smaller study that gives inaccurate results is precisely wasting money.
Actually, if one truly understands statistics, it turns out that obtaining a good sample is probably the most difficult task. If the sample truly represents the data points being studied, then 196 could very well be enough for useful information. Clearly there are a lot of thngs to worry about in a statistical study - the SIZE of the sample is one of those, but certainly not the most worrisome.
Besides, especially given the fact that in our society we have lots of Republicans, it might well be a legitimate finding to have "ten retards" or more in the sample.
If you go through the abstract the sample isn't very good either, only 42% were women and they also didn't break the sample down by race either.
|
