Monarchs and Monoculture in southern Michigan

[Stan Gorodenski]

Neil Jones wrote:

>No I wasn't grasping at straws. I was defeating a poor scientific argument 
>through standard scientific means, but you didn't realise that did you? The 
>trouble is you just looked at the figures. I looked at what the figures meant 
>and looked at what changing them slightly would do to the results. 
>The data didn't support the argument well. What did you expect? I think the 
>problem is that you believed it did. Try putting your figures through some 
>statistical tests. 
>
>  
>

To answer your question, the statistical test would probably show _no_ 
statistically significant difference before and after the treatment. 
However, in this situation it would be because of the large error term 
resulting from a small sample size that you alluded to. On the other 
hand, almost any effect, no matter how small, can be shown to be 
statistically significant if the sample is made large enough. The 
question then becomes - is the effect significant (not in a statistical 
sense)? I, and probably many others, including Ken, undoubtedly saw that 
the sample size was too small to come to any definite conclusions on 
this issue. However, irrespective of the shortcomings of the data that 
Ed (no Paul, if I understand the sequence of messages correctly) 
provided, it is good data because it is causing us to ask more questions 
and, obviously, to get more data to make more definitive statements at 
accepted significance levels. It is obvious that Paul was too 
enthusiastic in his response to the data Ed posted, given the small 
numbers involved. On the other hand, equally as bad, in my opinion, is 
someone coming out strongly on the other side through the technique of 
name calling, revealing facts unrelated to the discussion such as 
someone being banned from another list or the results of a personality 
tests. Recently, I read a book review of the clash between Chandrasekhar 
and Eddington. The former ended up being one of the great stellar 
theorists of our time. When he came out with a theory that disputed 
accepted views at the time, Eddington lambasted Chandra's ideas as being 
"stellar buffoonery". The result was that Chandra, being relatively new 
from India and apparently very sensitive, got out of white dwarf 
research. Years later Eddington apologized, but it is this kind of 
behavior, that, in my opinion, retards scientific research and the quest 
for the truth.
Stan



 
 ------------------------------------------------------------ 

   For subscription and related information about LEPS-L visit:

   http://www.peabody.yale.edu/other/lepsl