The default report format is the Crosstab format, which gathers all three statistics for each sample and response together. The Crosstab format displays the responses on the top and the sample categories down the side, with multiple table elements together in each cell of the cross tabulation.
Crosstab Format
The Crosstab format has a transposed version, Crosstab Transposed, which is useful when there are a lot of response categories but not a lot of samples. Crosstab Transposed displays the responses down the side and the sample categories across the top, with multiple table elements together in each cell.
The Legend displays or hides the legend for the response column on the Share Chart.
Calculates the response means, using the numeric categories, or value scores. This is enabled for columns that use numeric codes, or for categories that have a Value Scores property. To make the Mean Score interpretable, you can assign specific value scores in the Column Info window with the Value Scores column property. For more information and an example, refer to Mean Score Example.
1.
Open the Car Poll.jmp sample data table.
2.
Select Analyze > Consumer Research > Categorical.
3.
Select country and click Responses on the Simple tab.
4.
Select marital status and click X, Grouping Category.
5.
Click OK.
6.
Select Test Response Homogeneity from the Categorical red triangle menu.
Test Response Homogeneity
1.
Open the Failure3Freq.jmp sample data table in the Quality Control folder.
2.
Select Analyze > Consumer Research > Categorical.
3.
Select all of the defect columns and click Response Frequencies on the Multiple tab.
4.
Select clean and click X, Grouping Category to compare the samples across the clean treatment variable.
5.
Select SampleSize and click Sample Size.
6.
Click OK.
7.
Select Test Each Response from the Categorical red triangle menu.
Test Each Response
For which defects are the rates significantly different across the clean treatments? The p-values show that oxide defect is the most significantly different, followed by contamination, then doping. The other defects are not significantly different with this amount of data.
For single responses, Cell Chisq displays the cell-by-cell composition of the Pearson chi-square overall, and also shows which cells have relatively more (red) or less (blue) than expected if they were the same across sample categories. The value shown is the p-value for the chi-square. The color is bright when they are significant, and grayer when less significant, denoting visually where the significant differences are.
1.
Open the Consumer Preferences.jmp sample data table.
2.
Select Analyze > Consumer Research > Categorical.
3.
Select I am working on my career and click Responses on the Simple tab.
4.
Select Age Group and click X, Grouping Category.
5.
Click OK.
6.
Select Crosstab Transposed from the red triangle menu.
7.
Select Cell Chisq from the red triangle menu.
Cell Chisq
The Conditional Association option is used to compute the conditional probability of one response given a different response. A table and color map of the conditional probabilities are given. This option is available only when the Unique occurrences within ID box is checked on the Categorical launch window. A common application of this analysis is when the responses represent adverse events (side effects) from a drug. The computations represent the conditional probability of one side effect given the presence of another side effect. For AdverseR.jmp, given the response in each row, Conditional Association shows the rate of also having the response in a column. Conditional Association only displays a few variables in the table due to size constraints.
Conditional Association
The Rater Agreement analysis answers the questions of how closely raters agree with one another and if the lack of agreement is symmetrical. For example, open Attribute Gauge.jmp. The Attribute Chart script runs the Variability Chart platform, which has a test for agreement among raters.
Agreement Comparisons
Launch the Categorical platform and designate the three raters (A, B, and C) as Rater Agreement responses on the Related tab on the launch window. In the resulting report, you have a similar test for agreement that is augmented by a symmetry test that the lack of agreement is symmetric.
Agreement Statistics
1.
Open the Presidential Elections.jmp sample data table.
2.
Select Analyze > Consumer Research > Categorical.
3.
Select 1980 Winner through 2012 Winner and click Repeated Measures on the Related tab.
4.
Select State and click X, Grouping Category.
5.
Click OK.
Repeated Measures
Compare Each Sample
Compare Each Cell (cut off after column CF)
Lowercase letters are also used for comparisons that are slightly less significant, according to Letter Comparisons. These comparisons suffer when the count for that sample group (the Base Count) is small, and asterisks start to appear in the comparison cells to warn you.
The options are initialized to the current state. Select the appropriate options and select either Submit Platform Preferences or Create Platform Preference Script to submit the options to your preferences as the new default. When the Categorical platform is launched, the preferences associated with the current preference set are enacted.
Set Preferences Window
Free Text is used for comment fields where the analysis counts the frequency of each word used. Free Text gives word counts in both word order and frequency order, and the rate of non-empty text. The following example uses the Consumer Preferences.jmp sample data table, which contains survey data relating to oral hygiene preferences. A comment field was included in the survey asking for reasons why the participant did not floss.
1.
Open the Consumer Preferences.jmp sample data table, located in the Quality Control folder.
2.
Select Analyze > Consumer Research > Categorical.
3.
Select Reasons Not to Floss and click Free Text on the Multiple tab.
4.
Click OK.
5.
Select Score Words by Column from the red triangle menu and then select Floss. Click OK.
Free Text Report Example
Free Text Report Example details the free text word counts the respondents included as reasons why they do not floss. From the analysis, you can determine the number of words, cases, non-empty cases, and portions of non-empty cases. You can also view the word counts alphabetically, in terms of frequency, or by the mean scores. There are more commands to further customize the analysis on the Free Text red triangle menu on the report:
Treemap Example
The Structured tab enables you to construct complex tables of descriptive statistics by dragging column names into green icon drop zones to create side-by-side and nested results. The following example uses the Consumer Preferences.jmp sample data table. From this data, suppose that you wanted to compare job satisfaction and salary against gender by age group and position tenure.
1.
Open the Consumer Preferences.jmp sample data table.
2.
Select Analyze > Consumer Research > Categorical.
4.
Drag Gender to the green drop zone at the Top of the table on the Structured tab.
5.
Drag Age Group to the green drop zone just below Gender.
6.
Drag Position Tenure to the green drop zone at the Top of the table next to Gender.
7.
Drag Job Satisfaction to the green drop zone at the Side of the table.
8.
Drag Salary Group to the green drop zone at the Side of the table under Job Satisfaction.
Structured Tab Report Setup
9.
Click Add=>.
10.
Click OK.
Structured Tab Report Example
Structured Tab Report Example shows that the majority of both the male and female respondents were somewhat satisfied with their jobs, with the highest percentage of males being in the 25-29 age group, while the females were in the 30-34 age group. Most of those who were somewhat satisfied had been in their current position for less than 5 years.