Mosaic Plot	A graphical representation of the data in the Contingency Table. See “Mosaic Plot”.
Contingency Table	A two-way frequency table. There is a row for each factor level and a column for each response level. See “Contingency Table”.
Tests	Analogous to the Analysis of Variance table for continuous data. The tests show that the response level rates are the same across X levels. See “Tests”.
Set α level	Changes the alpha level used in confidence intervals. Select one of the common values (0.10, 0.05, 0.01) or select a specific value using the Other option.
Analysis of Means for Proportions	Only appears if the response has exactly two levels. Compares response proportions for the X levels to the overall response proportion. See “Analysis of Means for Proportions”.
Correspondence Analysis	Shows which rows or columns of a frequency table have similar patterns of counts. In the correspondence analysis plot, there is a point for each row and for each column of the contingency table. See “Correspondence Analysis”.
Cochran Mantel Haenszel	Tests if there is a relationship between two categorical variables after blocking across a third classification. See “Cochran-Mantel-Haenszel Test”.
Agreement Statistic	Only appears when both the X and Y variables have the same levels. Displays the Kappa statistic (Agresti 1990), its standard error, confidence interval, hypothesis test, and Bowker’s test of symmetry, also know as McNemar's test. See “Agreement Statistic”.
Relative Risk	Calculates risk ratios. Appears only when both the X and Y variables have only two levels. See “Relative Risk”.
Odds Ratio	Appears only when there are exactly two levels for each variable. Produces a report of the odds ratio. For more information, see “Statistical Details for the Odds Ratio Option”. The report also gives a confidence interval for this ratio. You can change the alpha level using the Set α Level option.
Two Sample Test for Proportions	Performs a two-sample test for proportions. This test compares the proportions of the Y variable between the two levels of the X variable. Appears only when both the X and Y variables have only two levels. See “Two Sample Test for Proportions”.
Measures of Association	Describes the association between the variables in the contingency table. See “Measures of Association”.
Cochran Armitage Trend Test	Tests for trends in binomial proportions across levels of a single variable. This test is appropriate only when one variable has two levels and the other variable is ordinal. See “Cochran Armitage Trend Test”.
Exact Test	Provides exact versions of the following tests: • Fisher’s Test • Cochran Armitage Trend Test • Agreement Test See “Exact Test”.

Table 7.1 Context Menu Options 

Set Colors

Shows the current assignment of colors to levels. See “Set Colors”.

Cell Labeling

Specify a label to be drawn in the mosaic plot. Select one of the following options: Unlabeled Shows no labels, and removes any of the other options. Show Counts Shows the number of observations in each cell. Show Percents Shows the percent of observations in each cell. Show Labels Shows the levels of the Y variable corresponding to each cell. Show Row Labels Shows the row labels for all of the rows represented by the cell.

Table 7.2 Description of the Select Colors for Values Window 

Macros	Computes a color gradient between any two levels, as follows: • If you select a range of levels (by dragging the mouse over the levels that you want to select, or pressing the SHIFT key and clicking the first and last level), the Gradient Between Selected Points option applies a color gradient to the levels that you have selected. • The Gradient Between Ends option applies a gradient to all levels of the variable. • Undo any of your changes by selecting Revert to Old Colors.
Color Theme	Changes the colors for each value based on a color theme.
Save Colors to Column	If you select this check box, a new column property (Value Colors) is added to the column in the associated data table. To edit this property from the data table, select Cols > Column Info.

Number	Description	Label in Table
36	Number of cars that are both large and come from America	Count
11.88%	Percentage of all cars that are both large and come from America (36/303)1.	Total%
85.71%	Percentage of large cars that come from America (36/42)2	Col%
31.30%	Percentage of American cars that are large (36/115)3.	Row%
37.95%	Percentage of all cars that come from America (115/303).	(none)
13.86%	Percentage of all cars that are large (42/303).	(none)

---

Table 7.3 Description of the Contingency Table 

Count	Cell frequency, margin total frequencies, and grand total (total sample size).
Total%	Percent of cell counts and margin totals to the grand total.
Row%	Percent of each cell count to its row total.
Col%	Percent of each cell count to its column total.
Expected	Expected frequency (E) of each cell under the assumption of independence. Computed as the product of the corresponding row total and column total divided by the grand total.
Deviation	Observed cell frequency (O) minus the expected cell frequency (E).
Cell Chi Square	Chi-square values computed for each cell as (O - E)2 / E.
Col Cum	Cumulative column total.
Col Cum%	Cumulative column percentage.
Row Cum	Cumulative row total.
Row Cum%	Cumulative row percentage.

Table 7.4 Description of the Tests Report 

N	Total number of observations.
DF	Records the degrees of freedom associated with the test. The degrees of freedom are equal to (c - 1)(r - 1), where c is the number of columns and r is the number of rows.
-LogLike	Negative log-likelihood, which measures fit and uncertainty (much like sums of squares in continuous response situations).
Rsquare (U)	Portion of the total uncertainty attributed to the model fit. • An R2 of 1 means that the factors completely predict the categorical response. • An R2 of 0 means that there is no gain from using the model instead of fixed background response rates. For more information, see “Statistical Details for the Tests Report”.
Test	Lists two Chi-square statistical tests of the hypothesis that the response rates are the same in each sample category. For more information, see “Statistical Details for the Tests Report”.
Prob>ChiSq	Lists the probability of obtaining, by chance alone, a Chi-square value greater than the one computed if no relationship exists between the response and factor. If both variables have only two levels, Fisher’s exact probabilities for the one-tailed tests and the two-tailed test also appear.

Table 7.5 Descriptions of the Analysis of Means for Proportions Options 

Set Alpha Level	Selects the alpha level used in the analysis.
Show Summary Report	Produces a report that shows the response proportions with decision limits for each level of the X variable. The report indicates whether a limit has been exceeded.
Switch Response Level for Proportion	Changes the response category used in the analysis.
Display Options	Shows or hides the decision limits, decision limit shading, center line, and needles.

Table 7.6 Descriptions of the Correspondence Analysis Options 

3D Correspondence Analysis	Produces a 3-D scatterplot.
Save Value Ordering	Takes the order of the levels sorted by the first correspondence score coefficient and makes a column property for both the X and Y columns.

Table 7.7 Description of the Details Report 

Singular Value	Provides the singular value decomposition of the contingency table. For the formula, see “Statistical Details for the Details Report”.
Inertia	Lists the square of the singular values, reflecting the relative variation accounted for in the canonical dimensions.
Portion	Portion of inertia with respect to the total inertia.
Cumulative	Shows the cumulative portion of inertia. If the first two singular values capture the bulk of the inertia, then the 2-D correspondence analysis plot is sufficient to show the relationships in the table.
X variable (Cheese) c1, c2, c3	The values plotted on the Correspondence Analysis plot.
Y variable (Response) c1, c2, c3	The values plotted on the Correspondence Analysis plot.

Table 7.8 Description of the Cochran-Mantel-Haenszel Test 

Correlation of Scores	Applicable when either the Y or X is ordinal. The alternative hypothesis is that there is a linear association between Y and X in at least one level of the blocking variable.
Row Score by Col Categories	Applicable when Y is ordinal or interval. The alternative hypothesis is that, for at least one level of the blocking variable, the mean scores of the r rows are unequal.
Col Score by Row Categories	Applicable when X is ordinal or interval. The alternative hypothesis is that, for at least one level of the blocking variable, the mean scores of the c columns are unequal.
General Assoc. of Categories	Tests that for at least one level of the blocking variable, there is some type of association between X and Y.

Table 7.9 Description of the Agreement Statistic Report 

Kappa	Shows the Kappa statistic.
Std Err	Shows the standard error of the Kappa statistic.
Lower 95%	Shows the lower endpoint of the confidence interval for Kappa.
Upper 95%	Shows the upper endpoint of the confidence interval for Kappa.
Prob>Z	Shows the p-value for a one-sided test for Kappa. The null hypothesis tests if Kappa equals zero.
Prob>|Z|	Shows the p-value for a two-sided test for Kappa.
ChiSquare	Shows the test statistic for Bowker’s test. For Bowker’s test of symmetry, the null hypothesis is that the probabilities in the square table satisfy symmetry, or that pij=pji for all pairs of table cells. When both X and Y have two levels, this test is equal to McNemar’s test.
Prob>ChiSq	Shows the p-value for the Bowker’s test.

Table 7.10 Description of the Two Sample Test for Proportions Report 

Description	Shows the test being performed.
Proportion Difference	Shows the difference in the proportions between the levels of the X variable.
Lower 95%	Shows the lower endpoint of the confidence interval for the difference. Based on the adjusted Wald confidence interval.
Upper 95%	Shows the upper endpoint of the confidence interval for the difference. Based on the adjusted Wald confidence interval.
Adjusted Wald Test	Shows two-tailed and one-tailed tests.
Prob	Shows the p-values for the tests.
Response <variable> category of interest	Select which response level to use in the test.

Table 7.11 Description of the Measures of Association Report 

Gamma	Based on the number of concordant and discordant pairs and ignores tied pairs. Takes values in the range -1 to 1.
Kendall’s Tau-b	Similar to Gamma and uses a correction for ties. Takes values in the range -1 to 1.
Stuart’s Tau-c	Similar to Gamma and uses an adjustment for table size and a correction for ties. Takes values in the range -1 to 1.
Somers’ D	An asymmetric modification of Tau-b. • The C|R denotes that the row variable X is regarded as an independent variable and the column variable Y is regarded as dependent. • Similarly, the R|C denotes that the column variable Y is regarded as an independent variable and the row variable X is dependent. Somers’ D differs from Tau-b in that it uses a correction for ties only when the pair is tied on the independent variable. It takes values in the range -1 to 1.
Lambda Asymmetric	• For C|R, is interpreted as the probable improvement in predicting the column variable Y given knowledge of the row variable X. • For R|C, is interpreted as the probable improvement in predicting the row variable X given knowledge about the column variable Y. Takes values in the range 0 to 1.
Lambda Symmetric	Loosely interpreted as the average of the two Lambda Asymmetric measures. Takes values in the range 0 to 1.
Uncertainty Coef	• For C|R, is the proportion of uncertainty in the column variable Y that is explained by the row variable X. • For R|C, is interpreted as the proportion of uncertainty in the row variable X that is explained by the column variable Y. Takes values in the range 0 to 1.
Uncertainty Coef Symmetric	Symmetric version of the two Uncertainty Coef measures. Takes values in the range 0 to 1.

Fisher’s Exact Test	Performs Fisher’s Exact test for an r x c table. This is a test for association between two variables. Fisher’s exact test assumes that the row and column totals are fixed, and uses the hypergeometric distribution to compute probabilities. This test does not depend on any large-sample distribution assumptions. This means it is appropriate for situations where the Likelihood Ratio and Pearson tests become less reliable, like for small sample sizes or sparse tables. The report includes the following information: Table Probability (P) gives the probability for the observed table. This is not the p-value for the test. Two-sided Prob ≤ P gives the p-value for the two-sided test. For 2x2 tables, the Fisher’s Exact test is automatically performed. See “Tests”.
Exact Cochran Armitage Trend Test	Performs the exact version of the Cochran Armitage Trend Test. This test is available only when one of the variables has two levels. For more details about the trend test, see “Cochran Armitage Trend Test”.
Exact Agreement Test	Performs an exact test for testing agreement between variables. This is an exact test for the Kappa statistic. This is available only when the two variables have the same levels. For more details about agreement testing, see “Agreement Statistic”.