Laserfiche WebLink
<br />Sta.nda1od ErTel' <br /> <br />Student's t <br />l)i.stzoibution <br />(t-distributionJ <br /> <br />Test of <br />Signifioanoe <br /> <br />Transformation <br /> <br />Val"ianoe <br /> <br />Weighted Meane <br /> <br />An estimate of the standard deviation of a sta- <br />tistic. Often calculated from a single set of <br />observations. Calculated like the standard devi- <br />ation but differing from it in meaning. <br /> <br />A distribution used in evaluation of variables <br />which involve sample standard deviation rather <br />than population standard deviation. <br /> <br />A test made to learn the probability that a <br />result is accidental or that a result differs from <br />another result. For all the many types of tests <br />there are standard formulas and tables. In making <br />a test it is necessary to choose a "level of signifi- <br />cance," the choice being arbitrary but generally not <br />less than the low level of 10 percent nor more than <br />the high level of 1 percent. <br /> <br />The change of numerical values of data to make <br />later computations easier, to linearize a plot <br />or to normalize a skewed distribution by making it <br />more nearly a normal distribution. The most common <br />transformations are those changing ordinary numerical <br />values into their logarithms, square roots or cube <br />roots; many others are possible. <br /> <br />A measure of the amount of spread or dispersion <br />of a set of values around their mean, obtained by <br />calculating the mean value of the squares of the <br />deviations from the mean, and hence equal to the <br />square of the standard deviation. <br /> <br />A value obtained by multiplying each of a series <br />of values by its assigned weight and dividing the <br />sum of those products by the sum of the weights. <br /> <br />2-5 <br />