The TwoSampleFTest function computes the two sample F-test upon datasets X1 and X2.  This calculation is used to determine the significance of the difference between the ratio of sample variances and an assumed ratio of population variances.

The first parameter X1 is the first data sample to use in the analysis.

The second parameter X2 is the second data sample to use in the analysis.

The third parameter beta is the assumed ratio of population variances (assumed population variance of X1 divided by the assumed population variance of X2), specified as a real constant.

The options argument can contain one or more of the options shown below.

alternative='twotailed', 'lowertail', or 'uppertail'

This option is used to specify the type or interval used in the analysis, or similarly, the alternative hypothesis to consider when performing the analysis.

confidence=float

This option is used to specify the confidence level of the interval and must be a floating-point value between 0 and 1.  By default this is set to 0.95.

ignore=truefalse

This option is used to specify how to handle non-numeric data. If ignore is set to true all non-numeric items in data will be ignored.

output='report', 'statistic', 'pvalue', 'confidenceinterval', 'distribution', 'hypothesis', or list('statistic', 'pvalue', 'confidenceinterval', 'distribution', 'hypothesis')

This option is used to specify the desired format of the output from the function.  If 'report' is specified then a module containing all output from this test is returned.  If a single parameter name is specified other than 'report' then that quantity alone is returned.  If a list of parameter names is specified then a list containing those quantities in the specified order will be returned.

weights1=rtable

Vector of weights (one-dimensional rtable). If these weights are given, the TwoSampleFTest function will weight each data point in X1 accordingly. Note that the weights provided must have type realcons and the results are floating-point, even if the problem is specified with exact values. Both the data array and the weights array must have the same number of elements.

weights2=rtable

Vector of weights (one-dimensional rtable).  This parameter is equivalent to the option weights1, except applying to data in X2.

This test generates a complete report of all calculations in the form of a userinfo message.  In order to access this report, specify infolevel[Statistics] := 1.

The chisquare test should be used to compare the variance (or standard deviation) of one sample against an assumed population variance (or standard deviation).

F-Ratio Test on Two Samples
---------------------------
Null Hypothesis:
Sample drawn from populations with ratio of variances equal to 1
Alt. Hypothesis:
Sample drawn from population with ratio of variances not equal to 1

Sample sizes:            10, 10
Sample variances:        18.0444, 12.1778
Ratio of variances:      1.48175
Distribution:            FRatio(9,9)
Computed statistic:      1.48175
Computed pvalue:         0.567368
Confidence interval:     .368046193452367 .. 5.96552419074047
                         (ratio of population variances)

Result: [Accepted]
There is no statistical evidence against the null hypothesis

F-Ratio Test on Two Samples
---------------------------
Null Hypothesis:
Sample drawn from populations with ratio of variances greater than 1
Alt. Hypothesis:
Sample drawn from population with ratio of variances less than 1

Sample sizes:            10, 10
Sample variances:        18.0444, 12.1778
Ratio of variances:      1.48175
Distribution:            FRatio(9,9)
Computed statistic:      1.48175
Computed pvalue:         0.716316
Confidence interval:     0 .. 4.71033065843087
                         (ratio of population variances)

Result: [Accepted]
There is no statistical evidence against the null hypothesis

F-Ratio Test on Two Samples
---------------------------
Null Hypothesis:
Sample drawn from populations with ratio of variances less than 1
Alt. Hypothesis:
Sample drawn from population with ratio of variances greater than 1

Sample sizes:            10, 10
Sample variances:        18.0444, 12.1778
Ratio of variances:      1.48175
Distribution:            FRatio(9,9)
Computed statistic:      1.48175
Computed pvalue:         0.283684
Confidence interval:     .4661219412315 .. infinity
                         (ratio of population variances)

Result: [Accepted]
There is no statistical evidence against the null hypothesis

Kanji, Gopal K. 100 Statistical Tests. London: SAGE Publications Ltd., 1994.

Sheskin, David J. Handbook of Parametric and Nonparametric Statistical Procedures. London: CRC Press, 1997.