StatsWRCorrelationTest

StatsWRCorrelationTest [/ALPH=val /T=k /Z/Q] waveA, waveB

The StatsWRCorrelationTest operation performs a Weighted Rank Correlation test on waveA and waveB, which contain the ranks of sequential factors. The waves are 1-based, integer ranks of factors in the range 1-2^31.

StatsWRCorrelationTest computes a top-down correlation coefficient using Savage sums as well as the critical and P-values. Output is to the W_StatsWRCorrelationTest wave in the current data folder or optionally to a table.

Flags

/ALPH=val	Sets the significance level (default val =0.05).
/DEST=dstWave	Specify the destination wave for the Weighted Rank Correlation test results. If you do not use this flag, the results are stored in the wave W_StatsWRCorrelationTest in the current data folder.
	This flag was added in Igor Pro 10.00.
/FREE	Creates the user-specified destination wave as a free wave.
	/FREE is allowed only in functions, and only if the destination waves are simple names or wave reference structure fields.
	See Free Waves for more discussion.
	The /FREE flag was added in Igor Pro 10.00.
/Q	No results printed in the history area.
/T=k	Displays results in a table. k specifies the table behavior when it is closed.
	k =0: Normal with dialog (default). k =1: Kills with no dialog. k =2: Disables killing.
	The table is associated with the test, not the data. If you repeat the test, it will update any existing table with the new results.
/Z	Ignores errors. V_flag will be set to -1 for any error and to zero otherwise.

Details

The StatsWRCorrelationTest input waves must be one-dimensional and have the same length. The waves are 1-based, integer ranks of factors corresponding to the point number. Ranks may have ties in which case you should repeat the rank value. For example, if the second and third entries have the same rank you should enter {1,2,2,4}. H₀ stipulates that the same factors are most important in both groups represented by waveA and waveB.

The top-down correlation is the sum of the product of Savage sums for each row:

$$r_{T D}=\frac{\displaystyle \sum_{i=1}^{n} S_{i A} S_{i B}-n}{n-S_{1}},$$

where n is the number of rows and the Savage sum S_i is

$$\displaystyle S_{i}=\sum_{j=i}^{n} \frac{1}{j},$$

and S_iA corresponds to the S_i value of the rank of the data in row (i -1) of waveA.

References

Iman, R.L., and W.J. Conover, A measure of top-down correlation, Technometrics, 29, 351-357, 1987.

See, in particular, Chapter 19 of:

Zar, J.H., Biostatistical Analysis, 4th ed., 929 pp., Prentice Hall, Englewood Cliffs, New Jersey, 1999.

See Also

Statistical Analysis, StatsLinearCorrelationTest, StatsRankCorrelationTest, StatsTopDownCDF, StatsInvTopDownCDF

Demos

Open WR Correlation Demo.pxp