Gossett

     Gossett is a statistical tool which allows a probability statement regarding whether or not

two independently selected samples represent a single population i.e. do the means differ.

     Student's t is used to determine the significance of difference between the means of the two

samples.

     t is then integrated to determine the probability of making a Type 1 error.

     A Type 1 error is the probability of error in rejecting the null hypothesis that the two

sample means come from the same population. This value is denoted as alpha.

(1-alpha) is then used as degree of confidence in stating that the two means differ.

     Gossett is used to select two(or more) variables from the MicroBase database for testing.

     Note that although a table is produced which presents the confidence level for all pairs

the results are only valid for each individual pair. No conclusion should be made about

a difference between all pairs. ANOVA is used to compare differences among many sample groups.

<EOF>

 

      

     