Details and restrictions of the cell chi-square test
Cell chi-square validity
The following conditions must be met in order for the cell chi-square test to be valid:
0 < column base < overall base
0 < row base < overall base
This means the test is not valid for a cell where all or no respondents belong in the whole column or row for the cell.
Significance level
You can select an explicit significance level of 10%, 5%, 2.5% 0.5% or 0.1%. Alternatively, an automatic selection can be requested, which sets the significance level depending on the overall base as follows:
|
Level applied
|
For overall base value “f” of
|
|
10%
|
f <= 300
|
|
5%
|
300 < f <= 1000
|
|
2.5%
|
1000 < f <= 4000
|
|
0.5%
|
4000 < f <= 20000
|
|
0.1%
|
f > 20000
|
Cell chi-square usage
The cell chi-square test is not suitable for all tables. When you request the test on a table that is structurally unsuitable, UNICOM Intelligence Professional omits the test, leaves the cell chi-square and p value rows blank, and writes an explanation to the diagnostics data. You must to make sure that the data in the table is generally suitable for testing, that the sample size is suitable, and so on.
UNICOM Intelligence Reporter displays a message if you define a cell chi-square test on an unsuitable table or if you change a table that has a cell chi-square test defined so that it is no longer suitable for the test. When this happens, you can either change the table so that it conforms to the restrictions, or remove the test from the table.
Hierarchical data
This test is unsuitable for running on lower level data when you are working with hierarchical data a hierarchical view of the data. See
Hierarchical data for more information.
Rows and columns
For the cell chi-square test, the variables on the side and top axes must have at least two categories. UNICOM Intelligence Professional does not perform the test on rows and columns in which all of the values are zero or on rows and columns that are formed from non-category elements, such as bases and means.
Nested and concatenated tables
When variables are concatenated in UNICOM Intelligence Professional, a base is inserted at the start of each variable (variables can, in general, be different). With nesting, a base is needed for each “nest section” (they are almost always different). When a cell has the cell chi-square test applied to it, the base row and column used will be the nearest base in each dimension. This is consistent with the base used for percentages and other statistics in UNICOM Intelligence Professional. When applied to the automatic selection option for significance level, this means that different parts of one table could have different significance levels applied.
Excluded Elements
The IncludeInBase=False property has no effect on the cell chi-square test. If a cell chi-square test is carried out on a table containing categories that are excluded from the base using IncludeInBase=False, the calculation includes rows and columns corresponding to the excluded categories.
See also