Page 38 - Quantitative Data Analysis
P. 38

Quantitative Data Analysis
                                              Simply Explained Using SPSS


                                          Outliers

               An outlier is an “extreme” value that does not fit in normal range of
               values. Outlier is an influential data point that can dramatically
               change the mean value.

               1.      Lower bound for an outlier:

               1.      A value in the data falling below the lower bound is
                       considered an outlier.

               2.      Upper bound for an outlier:

               3.      A value in the data falling above the upper bound is
                       considered an outlier.

               If a data set has one or more outliers, the marker of the maximum
               (or minimum) on the box-and-whisker plot is replaced by the    (or
                 ) and a dot is placed at the outliers.

               With reference to the data set in last example,
                               ;



                       IQR:













               Thus any value in the data set below 16.5 or above 30.50 will be
               considered as an outlier. Because no values are below 16.50 or
               above 30.50, this dataset has no outliers.





               The Theory and Applications of Statistical Inferences           22
   33   34   35   36   37   38   39   40   41   42   43