The R Project SVN R

Rev

Rev 88751 | Show entire file | Ignore whitespace | Details | Blame | Last modification | View Log | RSS feed

Rev 88751 Rev 89404
Line 55... Line 55...
55 
  (\eqn{n\equiv 2 \bmod 4}{n = 2 mod 4}), and are in the middle of
 55 
  (\eqn{n\equiv 2 \bmod 4}{n = 2 mod 4}), and are in the middle of
56 
  two observations otherwise.
 56 
  two observations otherwise.
57 
 
 57 
 
58 
  The notches (if requested) extend to \code{+/-1.58 IQR/sqrt(n)}.
 58 
  The notches (if requested) extend to \code{+/-1.58 IQR/sqrt(n)}.
59 
  This seems to be based on the same calculations as the formula with 1.57 in
 59 
  This seems to be based on the same calculations as the formula with 1.57 in
60 
  \bibcitet{|R:Chambers+Cleveland+Kleiner:1983|p.\\\\\\sspace{}62)}, given in
 60 
  \bibcitet{|R:Chambers+Cleveland+Kleiner:1983|p.\sspace{}62)}, given in
61 
  \bibcitet{|R:Mcgill+Tukey+Larsen:1978|p.\\\\\\sspace{}16)}.
 61 
  \bibcitet{|R:Mcgill+Tukey+Larsen:1978|p.\sspace{}16)}.
62 
  They are based on asymptotic normality of the median
 62 
  They are based on asymptotic normality of the median
63 
  and roughly equal sample sizes for the two medians being compared, and
 63 
  and roughly equal sample sizes for the two medians being compared, and
64 
  are said to be rather insensitive to the underlying distributions of
 64 
  are said to be rather insensitive to the underlying distributions of
65 
  the samples.  The idea appears to be to give roughly a 95\% confidence
 65 
  the samples.  The idea appears to be to give roughly a 95\% confidence
66 
  interval for the difference in two medians.
 66 
  interval for the difference in two medians.