NavList:

   

Reply

On 6/26/2015 3:26 PM, David Fleming wrote:
> I am going to repost my original data with errors of analysis, as
> pointed out by Bill B corrected and more extensive statistical analysis
> that is I think not too far from the truth.

After digging out some old stat books, I was way off the mark in
suggesting one could average the on- and off-the arc SDs. However, if
you do SD1^2 + SD2^2... you have not accounted for the different means,
although that is a lot of work for very little gain in this case.

(I still wonder whether on- vs off-the-arc are independent, as we are
measuring the diameter of the Sun with the same instrument in both
cases. it be cricket to combine them all as a string of 20, then do a
mean and variance/SD on the whole lot?)
>
> To me the bottom line baring mistakes is, as I originally maintained,
> overlap is supperior in measuring IC compared to edge to edge.

You cannot know that until you do an apple to apples comparison. From a
"safety net" vantage point, you have no idea if you are having a "bad
hair day" when superimposing other than a standard deviation. With
edge-to-edge you have standard deviations AND a reality check with the
semi-diameter or 4SD figures.

The above the brings us to the old question is the abbreviation "SD"
standard deviation or semi-diameter? (To which I believe Alex had a fix,
perhaps "s" for n-1 and "S" for n standard deviations?)

   

Reply