The ratio of the deviation, d, to the estimated standard deviation, s. It is used in experiments where the number of observations is small (31 or less and hence 30 or less degrees of freedom) and consequently the estimated standard deviation may differ widely from the true standard deviation. To find whether any deviation is significant, its probability is looked up in 'Student's' t tables, which take the number of degrees of freedom (N) into account. The greater the value of TV, the smaller the value of t for any given level of probability.
|