Displaying 1 result from an estimated 1 matches for "t312data".
Did you mean:
0x12data
2005 Apr 22
2
Hoaglin Outlier Method
...the wheel
or use a brute strength and force method if there is a better way in R.
Our usual method for determining outliers is a student's T test as in ASTM
E 178 or when the standard deviation for a lab is 3 or more. We normally
have 120 labs to evaluate for outliers similar what is shown in T312Data.
On occasion, I have used the Wilk-Shapiro W statistic in SAS. A point in
the right direction or an R code example would help greatly. After I trim
the outliers, I will need to show which labs were eliminated but that
should be fairly trivial.
The reference in Appendix A is:
Hoaglin, D. C., Igle...