Experimental Designs I Statistics Project Example | Topics and Well Written Essays - 750 words

Experimental Designs I - Statistics Project Example 4. It sets the rate of alpha error to the experiment error rate, which is usually 0.05. This is divided by comparisons in totality to type 1 error control if there is a consideration of multiple comparisons. In case, the bonferroni test is applied, there will be a print out of multiple comparison tables by SPSS providing mean differences in dependent variables among groups. The importance of these differences is also given showing 0.05 to be the differences significant level. 6. Post hoc comparisons are performed when a researcher is finding out differences, which is not limited to an individual’s theory (Gonzalez, 2008). Many tests that are carried out under post hoc apply the q statistics. If group means comparisons are chosen because of their large size, there is a variability increase expected. This must be compensated by the researcher through application of more tests, or else there will be occurrences of type 1 errors. 8. Repeated measures ANOVA is more powerful because every factor controls itself. In these designs, differences in individual subjects do not interfere with treatment group differences (Kulinskaya & Dollinger, 2007). SS stands for variation; df means the degree of freedom; MS is the variance that is arrived at by SS/df, and F is the ratio test got by dividing between MS by within MS. In the table above, the MS of within-group is less than between groups which shows that the grouping has no effect. The grouping has been done in three categories, that is, df within groups being (3-1) =2. There are 4 people in every group, therefore, df within groups is the group number multiplied by one less the number of each group: 3*2= 6. These are denominator’s and numerator’s df. From the F table with 0.0 5significance level, 5.14 is the critical value. As the F value that has been computed is less, it can be concluded that grouping variables have no effect on dependent variable. SS stands