Location: Commodity Utilization ResearchTitle: ANOVA for unbalanced data with missing cells: using spreadsheets to evaluate SAS Type IV sums of squares
Submitted to: American Chemical Society SciMeetings
Publication Type: Other
Publication Acceptance Date: 5/28/2021
Publication Date: 5/28/2021
Citation: Klasson, K.T. 2021. ANOVA for unbalanced data with missing cells: using spreadsheets to evaluate SAS Type IV sums of squares. American Chemical Society SciMeetings. https://doi.org/10.1021/scimeetings.1c00598.
Technical Abstract: Students and researchers are often introduced to Analysis of Variance (ANOVA) statistics by first studying one-way ANOVA examples. They then move on to two-way ANOVA, but often this is limited to balanced data. Balanced data are when the number of data values for each category (or group) is the same. When this is the case, the calculations are easy, and the equations needed for an ANOVA table are simple. When unbalanced data are encountered, which is often the case in agricultural field studies, the data are often entered into a statistical program such as SAS to construct the ANOVA table. The most complicated situation is when data are completely missing in a two-way ANOVA design for some treatment combination(s). This can be handled by a SAS Type IV evaluation in the SAS GLM Procedure. In this presentation we will show how dynamic spreadsheets can be constructed to produce SAS Type IV Sums of Squares for two-way ANOVA tables using hypotheses testing. Also, different hypotheses can be constructed by the user.