NAG Library Routine Document
G08AHF performs the Mann–Whitney test on two independent samples of possibly unequal size.
|SUBROUTINE G08AHF (
||N1, X, N2, Y, TAIL, U, UNOR, P, TIES, RANKS, WRK, IFAIL)
||N1, N2, IFAIL
||X(N1), Y(N2), U, UNOR, P, RANKS(N1+N2), WRK(N1+N2)
The Mann–Whitney test investigates the difference between two populations defined by the distribution functions and respectively. The data consist of two independent samples of size and , denoted by and , taken from the two populations.
The hypothesis under test,
, often called the null hypothesis, is that the two distributions are the same, that is
, and this is to be tested against an alternative hypothesis
- : ; or
- : , i.e., the 's tend to be greater than the 's; or
- : , i.e., the 's tend to be less than the 's,
using a two tailed, upper tailed or lower tailed probability respectively. You select the alternative hypothesis by choosing the appropriate tail probability to be computed (see the description of parameter TAIL
in Section 5
Note that when using this test to test for differences in the distributions one is primarily detecting differences in the location of the two distributions. That is to say, if we reject the null hypothesis in favour of the alternative hypothesis : we have evidence to suggest that the location, of the distribution defined by , is less than the location, of the distribution defined by .
test differs from the Median test (see G08ACF
) in that the ranking of the individual scores within the pooled sample is taken into account, rather than simply the position of a score relative to the median of the pooled sample. It is therefore a more powerful test if score differences are meaningful.
The test procedure involves ranking the pooled sample, average ranks being used for ties. Let
be the rank assigned to
the rank assigned to
. Then the test statistic
is defined as follows;
is also the number of times a score in the second sample precedes a score in the first sample (where we only count a half if a score in the second sample actually equals a score in the first sample).
||The test statistic .
||The approximate Normal test statistic,
is the number of groups of ties in the sample and is the number of ties in the th group.
Note that if no ties are present the variance of reduces to .
||An indicator as to whether ties were present in the pooled sample or not.
||The tail probability, , corresponding to (adjusted to allow the complement to be used in an upper one tailed or a two tailed test), depending on the choice of TAIL, i.e., the choice of alternative hypothesis, . The tail probability returned is an approximation of is based on an approximate Normal statistic corrected for continuity according to the tail specified. If and are not very large an exact probability may be desired. For the calculation of the exact probability see G08AJF (no ties in the pooled sample) or G08AKF (ties in the pooled sample).
The value of can be used to perform a significance test on the null hypothesis against the alternative hypothesis . Let be the size of the significance test (that is, is the probability of rejecting when is true). If then the null hypothesis is rejected. Typically might be or .
Conover W J (1980) Practical Nonparametric Statistics Wiley
Neumann N (1988) Some procedures for calculating the distributions of elementary nonparametric teststatistics Statistical Software Newsletter 14(3) 120–126
Siegel S (1956) Non-parametric Statistics for the Behavioral Sciences McGraw–Hill
- 1: N1 – INTEGERInput
On entry: the size of the first sample, .
- 2: X(N1) – REAL (KIND=nag_wp) arrayInput
On entry: the first vector of observations, .
- 3: N2 – INTEGERInput
On entry: the size of the second sample, .
- 4: Y(N2) – REAL (KIND=nag_wp) arrayInput
On entry: the second vector of observations. .
- 5: TAIL – CHARACTER(1)Input
: indicates the choice of tail probability, and hence the alternative hypothesis.
- A two tailed probability is calculated and the alternative hypothesis is .
- An upper tailed probability is calculated and the alternative hypothesis , i.e., the 's tend to be greater than the 's.
- A lower tailed probability is calculated and the alternative hypothesis , i.e., the 's tend to be less than the 's.
, or .
- 6: U – REAL (KIND=nag_wp)Output
On exit: the Mann–Whitney rank sum statistic, .
- 7: UNOR – REAL (KIND=nag_wp)Output
: the approximate Normal test statistic,
, as described in Section 3
- 8: P – REAL (KIND=nag_wp)Output
: the tail probability,
, as specified by the parameter TAIL
- 9: TIES – LOGICALOutput
: indicates whether the pooled sample contained ties or not. This will be useful in checking which routine to use should one wish to calculate an exact tail probability.
, no ties were present (use G08AJF
for an exact probability).
, ties were present (use G08AKF
for an exact probability).
- 10: RANKS() – REAL (KIND=nag_wp) arrayOutput
: contains the ranks of the pooled sample. The ranks of the first sample are contained in the first N1
elements and those of the second sample are contained in the next N2
- 11: WRK() – REAL (KIND=nag_wp) arrayWorkspace
- 12: IFAIL – INTEGERInput/Output
must be set to
. If you are unfamiliar with this parameter you should refer to Section 3.3
in the Essential Introduction for details.
For environments where it might be inappropriate to halt program execution when an error is detected, the value
is recommended. If the output of error messages is undesirable, then the value
is recommended. Otherwise, if you are not familiar with this parameter, the recommended value is
. When the value is used it is essential to test the value of IFAIL on exit.
unless the routine detects an error or a warning has been flagged (see Section 6
6 Error Indicators and Warnings
If on entry
, explanatory error messages are output on the current error message unit (as defined by X04AAF
Errors or warnings detected by the routine:
|On entry,||, or .|
The pooled sample values are all the same, that is the variance of .
The approximate tail probability, , returned by G08AHF is a good approximation to the exact probability for cases where and . The relative error of the approximation should be less than , for most cases falling in this range.
The time taken by G08AHF increases with and .
This example performs the Mann–Whitney test on two independent samples of sizes and respectively. This is used to test the null hypothesis that the distributions of the two populations from which the samples were taken are the same against the alternative hypothesis that the distributions are different. The test statistic, the approximate Normal statistic and the approximate two tail probability are printed. An exact tail probability is also calculated and printed depending on whether ties were found in the pooled sample or not.
9.1 Program Text
Program Text (g08ahfe.f90)
9.2 Program Data
Program Data (g08ahfe.d)
9.3 Program Results
Program Results (g08ahfe.r)