Steps for ANOVA 1. Find total variation (SST - Sum of squares total) 2. Find variation between/among levels (SSA- Sum of squares among levels or groups) 3. Find variation within levels/groups (SSW-Sum of squares within groups or levels SSW=SST-SSA Note: In the textbook, SSW is represented using SSE. 4. Find MSA (Mean squared variation between groups/levels) MSA=SSA/(c-1) Where c=total number of levels or groups you are comparing Note: In some books 'c' is denoted by the symbol 'a' or 'k', SSA is referred to as sum of squares treatments. SSA in your text book is represented as SSTR. MSA is also referred to as mean square treatments. 5. Find MSW (Mean squared variation within levels) MSW=SSW/(c(n-1)) n= number of observations in each group or level 6. Find F test statistic F=MSA/MSW Rejection rule: F>Falpha, c-1, c (n-1) Note: alpha =significance level (0.05, 0.1 or 0.01 if not stated in the problem) Degrees of Sum of Mean Source freedom Squares Square Within Treatment c(n-1) SSW MSW F MSA/MSW Between Treatment c-1 SSA MSA Hypothesis statements Ho: mu1=mu2=mu3=mu4 Ha: Not all hospital population means are equal Anova: Single Factor SUMMARY Groups Column 1 Column 2 Column 3 Column 4 Test Statistic Count Sum 4 4 4 4 ANOVA Source of Variation Between Groups Within Groups SS 61.25 222.5 Total 283.75 Conclusion: Since F(test statistic)