Hi all, when looking at LibreOffice bug 93329, I have looked at section 6.18.30 FTEST, ODF1.2 part 2. I found some problems. The file "FTEST Problems and Improvements" contains the problem description and a proposal for discussion. The file "FTest_With_Wizard.gnumeric" contains the result of the wizard. You can tell the wizard to write formulas. In the formulas you will find the use of "min" (see problem description). The file "UntersucheAlgorithm.ods" contains a cross-table with a lot of values for degrees of freedom. In case the test statistic has a sample value a little bit larger then 1, the right tail might become larger than 0.5, which results in a return value from FTEST larger than 1. The left, top cell contains a value which will be added to 1. Change it to see the effect in the table. Wrong values are colored by conditional formatting. Because FDIST calculates the right tail, a value smaller than 0.5 will result in a value larger 0.5, when calculating the left tail. Therefore in case the sample value of the test statistic is smaller than one, values smaller 0.5 are colored as errors. To have a test case with data sets, I have attached file "ErrorExampleForODF.ods" Kind regards Regina Attachment: FTEST Problems and Improvements.odt Description: application/vnd.oasis.opendocument.text Attachment: FTest_With_Wizard.gnumeric Description: application/gnumeric Attachment: UntersucheAlgorithm.ods Description: application/vnd.oasis.opendocument.spreadsheet Attachment: ErrorExampleForODF.ods Description: application/vnd.oasis.opendocument.spreadsheet