View Single Post
Old December 28, 2013, 12:49 PM   #2
Unclenick
Staff
 
Join Date: March 4, 2005
Location: Ohio
Posts: 21,017
Have you evaluated the mean points of impact (POI's)? If you know the horizontal and vertical differences between actual shot locations and the points of aim (POA's) used to fire them (this eliminates the aiming error you mentioned) it is easy to use Excel to spot trends in both group size and center location. Ideally you don't want the group centers to shift over a span of three or more load increments. Plotting a running 3 shot average of the POI quickly finds the best candidate. If you assemble the data and want help with this, let me know.

The reason for the above is overcoming the 3 shot group size's limited statistical significance. By itself, based on just the one three shot sample, any given 3 shot group predicts just over a 6:1 range of group sizes that fall within the 95% confidence limits for future groups. This goes from roughly 0.4 times smaller to 2.5 times bigger, meaning future groups are expected to be within that range 19 times out of 20, and outside that range (even smaller or even larger) 5% of the time. But by evaluating the average of three of your three-shot groups fired in a row (three charge weight steps in a row) as one group, you are looking at a collective 9 shot group, for which the 95% confidence limits are much tighter, with a smallest to largest ratio of about 1.5:1; from about 0.8 times smaller to less than 1.3 times bigger. It provides a much more trustworthy result. I assume it is the reason Newberry looks for three groups in a row with the same POI.
__________________
Gunsite Orange Hat Family Member
CMP Certified GSM Master Instructor
NRA Certified Rifle Instructor
NRA Benefactor Member and Golden Eagle
Unclenick is offline  
 
Page generated in 0.03945 seconds with 8 queries