ANOVA
A Quick Review
    Categorical: A variable with a discrete number of values; an ordinal or nominal variable. Categorical
                    variables are often used as grouping variables or factors.
    Factors: An independent variable defining groups of cases.
    Levels: The values of a factor are referred to as levels of the factor, or factor levels


To perform a Univariate Anova:  Remember the dependent variable should be continuous in nature,
                                                  and the grouping variables should be categorical
.
    Analyze
        General Linear Model
                Univariate

1. In the Dependent Box put over the variable that you are measuring (should be quantitative)

2. In the Factors box, move over the grouping variables, factors are categorical

3. In the Covariate box put quantitative variables that are related to the dependent variable. (advanced option)

4. The Plots button will let you graph the analysis, make sure you put one each of the grouping variables on the X-Axis, (Horizontal Axis)
            and add them then put one of the grouping factors on the X-axis (Horizontal Axis) and the other as separate lines and add that.

5. Click Options, Then Select: Descriptive Statistics, Estimates of Effect Size, and Observed Power
             If you want the St. Error or Confidence Intervals, move the grouping variables and/or the interaction
             to the left.

6. The output will give you 3 or 4 boxes
        1. The first box gives you the levels of the variables you asked for and their N

        2. The second box gives you the descriptive statistics for each of the specified levels:
                The first variable, labeled Dependent, is the actual values you will see, as well as the
                        Std. Deviations and N

                The first variable across the top is the first grouping variable.  The second variable is the second grouping variable. 
                        The means, Std. deviations, and N's are the dependent variable for each of these the different groups.

                The first grouping variable is split up into rows.  In the first row, the number on the left refers to the level of the first grouping variable,
                        the numbers on the right refer to the levels of the second grouping variable.  The Total refers to the all the different levels of the first group.

                The following rows are for the different levels of the first grouping variable.

                The very last row includes the totals for the second grouping variable.  The levels of that variable are defined on the right.
                        All levels of the first group are included in the totals.

         3. The third box called Tests of Between-Subjects Effects includes your F-values and df's.
                The first variable, labeled Dependent, is the actual values you will see.

                Rows:
                        Corrected Model- we ignore
                        Intercept row- we ignore
                        The next rows will be your grouping variables, followed by the interactions.
                        Error row- has the degrees of freedom that we use
                        Total- we ignore
                        Corrected Total row- we ignore
               
                Columns:
                      Type III Sums squares, we ignore
                      Degrees of Freedom is in the Error row.
                      Mean Square we ignore.
                      F- contains the F-Value that has been calculated
                      Sig- contains the Significance Level
                      Partial Eta Squared- is the Effect Size (.3 is small  .5 is medium   .8 is large, this value can go over 1.0)
                      Noncent Parameter.- we ignore this
                      Observed Power- The Power, 1.0 is the highest

          4. The last box, Estimated Marginal Means, will contain the Means, Std. Error and Confidence Intervals for the grouping
                   variables and the interaction, if you selected it under Options.



page created by Ryan Pohlig