This menu performs statistical tests on contingency tables.
Chi-square test
You can calculate the chi-square test of association between the factors
classifying the rows and the columns of a two-way tables of counts.
Method
You can choose between two ways of calculating the test: Pearson uses the
familiar method; the alternative, which may be more accurate, is to use maximum
likelihood.
Data Arrangement
For a chi-square test you can supply the data in one of three ways.
- Table - you can supply a two-way table containing the counts.
You can construct a table using the Create Table button, where you create a table
from existing data or by typing the values into a new spreadsheet.
- Row and Column Factors - you can supply 2 factors; one containing
rows and one containing columns. GenStat will automatically create a two-way table
of frequency counts and this will be used in the test.
- Single Variate with grouping factors - you can supply the counts in a variate
along with classifying factors for the rows and columns. This is the same layout as
you would use in a log-linear analysis. GenStat automatically forms a two-way table
from these data structures and performs a chi-square test on the table.
Fisher's exact test
This menu performs Fisher's exact test of association between the factors
classifying a 2x2 table of counts. It provides the one-tailed significance
level from the exact test, together with the mid-p value, which includes only
half the probability of the observed data table. Two methods are used to
calculate a two-tailed significance level. The first simply doubles the
one-tailed significance level, whereas the second method calculates the
cumulative probability of all outcomes that are no more probable than the
observed table. Mid-p values are produced for each of these methods.
Table (for Fisher's exact test)
Specifies the two-way, 2x2 table containing the counts.
You can construct a table using the Create Table button, where you create a table
from existing data or by typing the values into a new spreadsheet. An alternative
way to enter the data for a Fisher's exact test is to just enter the four values
from the 2x2 table into the Table box separating them by commas. For example:
4,6,7,6
Available Data
This lists data structures appropriate to the current input field. The contents
will change as you move from one field to the next. Double-click on a name to
copy it to the current input field; alternatively, you can type the name directly
into the input field.
See Also
- Procedure FEXACT2X2
is used to perform the calculations for Fisher's Exact Test. If you use this in command mode,
you can also set its PRINT option to 'tables' to display all 2x2 tables with margins that are
the same as the observed table, ogether with their probabilities of occurrence under the null hypothesis of
no association and the cumulative probabilities calculated from both tails.
- CHISQUARE procedure for calculating chi-square statistics
in command mode.
- Chi-square Goodness of fit
- Summaries of Groups (Tabulation) to create tables from existing data
- Contingency Table Options
- Contingency Table Save Options