Please note that the recommended version of Scilab is 2024.1.0. This page might be outdated.

See the recommended documentation of this function

# ftuneq

Fisher ratio for samples of unequal size.

### Syntax

f=ftuneq(sample1[,sample2[,sample3]...]]) [f,p]=ftuneq(sample1[,sample2[,sample3]...]])

### Arguments

- sample1, sample2, sample3,...
real or complex matrix of any type

### Description

This function computes the F ratio for samples of unequal size.

"The most efficient design is to make all samples the same size n. However when this is nor feasible, it still is possible to modify the ANOVA calculations."

Note that the definition of xbarbar is no longer
mean(xbar), but rather a weighted average with weights
ni. Additionally it gives (in p) the p-value of the
computed Fisher ratio. |

Given a number a of samples each of them composed of n_i (i from 1 to a) observations this function computes in f the Fisher ratio (it is the ratio between nr times the variance of the means of samples and the mean of the variances of each sample).

`f=ftest(samples)`

computes the Fisher ratio of the
nc samples whose values are in the columns of the matrix
`samples`

. Each one of these samples is composed of nr
values. (The Fisher ratio is the ratio between nr times
the variance of the means of samples and the mean of
variances of each sample)

`[f,p]=ftest(samples)`

gives in p the p-value of the
computed Fisher ratio f.

### References

Wonacott, T.H. & Wonacott, R.J.; Introductory Statistics, J.Wiley & Sons, 1990.

### Examples

sample1=[46 55 54]; sample2=[53 54]; sample3=[50 49 58 51 50]; sample4=[61 51 46 52]; [f,p]=ftuneq(sample1,sample2,sample3,sample4)

### See also

- ftuneq — Fisher ratio for samples of unequal size.

Report an issue | ||

<< ftest | Tests d'hypothèses | Matrices creuses >> |