Let the ordered sample observations be denoted byy1, y2, , yn (y1 being the smallest and yn the largest). Oursuggested check for normality is to plot the(F21((i 2 .5)yn), yi) pairs. Suppose we believe that theobservations come from a distribution with mean 0, and letw1,, wn be the ordered absolute values of the xi9s.Ahalfnormal plot is a probability plot of the wi9s. Morespecifically,since P(uZu # w) 5 P(2w # Z # w) 52F(w) 2 1, a half-normal plot is a plot of the(F21y{[(i 2 .5)yn 1 1]y2}, wi) pairs. The virtue of thisplot is that small or large outliers in the original sample willnow appear only at the upper end of the plot rather than at bothends. Construct a half-normal plot for the following sample ofmeasurement errors, and comment: 23.78, 21.27, 1.44,2.39, 12.38, 243.40, 1.15, 23.96, 22.34, 30.84

The Coefficient of Variation: C.V. = S / x(bar) X(bar) Mean of data Examples: #1 X(bar) 54 ft S 26 ft C.V. = S / x(bar) = 26/54 = .48 or 48% #2 X(bar) 2 hours S 2.7 hours C.V. = 2.7 / 2 = 1.35 NOTICE: the units cancel when you divide the two numbers, therefore the answer does not have units!! Chebyster’s Theorem: For any set of numbers with mean x(bar) and standard...