Test for Outliers
Outliers can occur by chance in any distribution, but they are often indicative either of measurement error or that the population has a heavy-tailed distribution. With Grubbs' test, an outlier is defined as as a value with Z>G and SD>SD0 where,
x = the value of the suspected outlier
Xgroup = group mean
N = number of measurements in the group
SD0 is the value you enter
SD and G are calculated for each group in each column (gene), and Z is calculated for each value. Without Grubb's, a group is considered to include an outlier if SD>SD0.
Select Test for Outliers in the Pre-processing menu in the Data editor to open a dialog box. Use the upper check box to decide whether to use Grubbs' test or not. Select the appropriate classification column in the drop down list, to be used to group values for correct mean and SD calculations. Choose the Confidence level and the cutoff SD (cycles) in the two text boxes, and whether to delete the found outliers in the bottom check box. If the check box is left unticked, the outliers will be marked in red in the data sheet.
