DAVID TEST

DAVID TEST

Analysis Command

Perform the David, Hartley and Pearson test for univariate outliers from a normal distribution.

The David, Hartley and Pearson statistic tests whether the minimum and maximum values from a univariate dataset are simultaneously outliers. This test assumes that the data come from an approximately normal distribution.

The test statistic is

\[ D = \frac{r} {s} \]

with \( r \) and \( s \) denoting the sample range and sample standard deviation, respectively.

Dataplot supports several methods for deterining the critical values for this test.

1. The ASTM E178-16a standard provides tables for n = 3 to 50 and alpha levels of 0.10, 0.05 and 0.01. Linear interpolation is used for values of n not given in the table. For values of n > 50, the simulation (see below) method is used.

2. The original paper by David contains tables for n = 3 to 1,000 and alpha levels of 0.10, 0.05, 0.025, 0.01 and 0.005. Linear interpolation is used for values of n not given in the table.

3. The David paper suggests the following formula (equation 6 on page 485)

   \[ cv = \sqrt{ \frac{2(n-1)t_{((1-\alpha)/(n(n-1)),n-2)}^{2}} {(n-2) + t_{((1-\alpha)/(n(n-1)),n-2)}^{2}} } \]

   with t denoting the percent point function of the t distribution.

4. Critical values can be obtained via simulation.

To specify the method used to compute the critical value, enter one of the following commands (the default is ASTM)

SET DAVID TEST CRITICAL VALUES ASTM
SET DAVID TEST CRITICAL VALUES DAVID
SET DAVID TEST CRITICAL VALUES FORMULA
SET DAVID TEST CRITICAL VALUES SIMULATION


This test is included in the ASTM E178 standard for outliers.

DAVID TEST <y> <SUBSET/EXCEPT/FOR qualification>
where <y> is the response variable being tested;
and where the <SUBSET/EXCEPT/FOR qualification> is optional.

MULTIPLE DAVID TEST <y1> ... <yk>
                        <SUBSET/EXCEPT/FOR qualification>
where <y1> ... <yk> is a list of up to k response variables;
and where the <SUBSET/EXCEPT/FOR qualification> is optional.

This syntax performs the David test on <y1>, then on <y2>, and so on. Up to 30 response variables may be specified.

Note that the syntax

MULTIPLE DAVID TEST Y1 TO Y4

is supported. This is equivalent to

MULTIPLE DAVID TEST Y1 Y2 Y3 Y4

REPLICATED DAVID TEST <y> <x1> ... <xk>
                        <SUBSET/EXCEPT/FOR qualification>
where <y> is the response variable;
            <x1> ... <xk> is a list of up to k group-id variables;
and where the <SUBSET/EXCEPT/FOR qualification> is optional.

This syntax performs a cross-tabulation of <x1> ... <xk> and performs a David test for each unique combination of cross-tabulated values. For example, if X1 has 3 levels and X2 has 2 levels, there will be a total of 6 David tests performed.

Up to six group-id variables can be specified.

Note that the syntax

REPLICATED DAVID TEST Y X1 TO X4

is supported. This is equivalent to

REPLICATED DAVID TEST Y X1 X2 X3 X4

DAVID TEST Y1
MULTIPLE DAVID TEST Y1 Y2 Y3
REPLICATED DAVID TEST Y X1 X2
DAVID TEST Y1 SUBSET TAG > 2

Tests for outliers are dependent on knowing the distribution of the data. The David test assumes that the data come from an approximately normal distribution. For this reason, it is strongly recommended that the David test be complemented with a normal probability test. If the data are not approximately normally distributed, then the David test may be detecting the non-normality of the data rather than the presence of an outlier.

You can specify the number of digits in the David output with the command
SET WRITE DECIMALS <value>

The DAVID TEST command automatically saves the following parameters:

STATVAL	=	the value of the test statistic
STATDCF	=	the CDF value of the test statistic
PVALUE	=	the p-value of the test statistic
CUTOFF80	=	the 80 percent point of the reference distribution
CUTOFF90	=	the 90 percent point of the reference distribution
CUTOFF95	=	the 95 percent point of the reference distribution
CUTOF975	=	the 97.5 percent point of the reference distribution
CUTOFF99	=	= the 99 percent point of the reference distribution

The STATCDF and PVALUE are only saved when the simulation method is used to obtain critical values. If the ASTM method is used to obtain critical values, the CUTOFF80 and CUTOF975 values are not saved. When the DAVID method is used to obtain critical values, the CUTOFF80 value is not saved.

If the MULTIPLE or REPLICATED option is used, these values will be written to the file "dpst1f.dat" instead.

In addition to the DAVID TEST command, the following commands can also be used:
LET A = DAVID TEST Y
LET A = DAVID TEST CDF Y
LET A = DAVID TEST PVALUE Y
LET A = DAVID TEST MINIMUM INDEX Y
LET A = DAVID TEST MAXIMUM INDEX Y


LET ALPHA = <value>
LET A = DAVID TEST CRITICAL VALUE Y


The DAVID TEST, DAVID TEST CDF, and DAVID TEST PVALUE return the values of the test statistic, the cdf of the test statistic and the pvalue of the test statistic, respectively. For the DAVID TEST CDF and DAVID TEST PVALUE commands, the simulation method will be used. Otherwise, the method specified by the SET DAVID TEST CRITICAL VALUE command will be used.

The DAVID TEST MINIMUM INDEX and DAVID TEST MAXIMUM INDEX return the row index of the minimum and maximum values of the response variable, respectively.

The DAVID TEST CRITICAL VALUE returns the critical value for the specified value of ALPHA. If ALPHA is not specified, it will be set to 0.05. Note that if the ASTM or DAVID methods are specified for the critical values, only a few select values for alpha are supported (0.01, 0.05 and 0.10 for ASTM and 0.005, 0.01, 0.025, 0.05 and 0.10 for DAVID).

In addition to the above LET command, built-in statistics are supported for about 25 different commands (enter HELP STATISTICS for details).

The ASTM method is used to obtain critical values

None

GRUBBS TEST	=	Perform the Grubbs outlier test.
TIETJEN-MOORE TEST	=	Perform a Tietjen-Moore outlier test.
EXTREME STUDENTIZED DEVIATE TEST	=	Perform a extreme studentized deviate outlier test.
DIXON TEST	=	Perform a Dixon outlier test.
SKEWNESS TEST	=	Perform the skewness outlier test.
KURTOSIS OUTLIER TEST	=	Perform the kurtosis outlier test.
GOODNESS OF FIT TEST	=	Perform a goodness of fit test (Anderson-Darling, Kolmogorov-Smirnov, chi-square, PPCC)
WILKS SHAPIRO NORMALITY TEST	=	Perform a Wilks Shapiro normality test.
HISTOGRAM	=	Generate a histogram.
PROBABILITY PLOT	=	Generates a probability plot.
BOX PLOT	=	Generate a box plot.

David, Hartley, and Pearson (1954), "The Distribution of the Ratio, in a Single Normal Sample, of Range to Standard Deviation", Biometrika, Vol. 41, pp. 482-493.

E178 - 16A (2016), "Standard Practice for Dealing with Outlying Observations", ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, USA.

Outlier Detection

2019/10

 
. Step 1:   Read the data - example from ASTM E178 standard
.
read y
-1.40
-0.44
-0.30
-0.24
-0.22
-0.13
-0.05
0.06
0.10
0.18
0.20
0.39
0.48
0.63
1.01
end of data
.
. Step 2:   Perform the DAVID TEST
.
set write decimals 3
set david test critical values astm
david test y
set david test critical values david
david test y
set david test critical values formula
david test y
set david test critical values simulation
david test y
    

The following output is generated

THE FORTRAN COMMON CHARACTER VARIABLE DAVITEST HAS JUST BEEN SET TO ASTM
 
            David, Hartley, Pearson Test for Outliers: Simultaneous Test
                  for Minimum and Maximum (Assumption: Normality)
 
Response Variable: Y
 
H0: The minimum and maximum are not
    both outliers
Ha: Both the minimum and maximum are
    outliers
Potential minimum outlier value tested:          -1.400
Potential maximum outlier value tested:           1.010
 
Summary Statistics:
Number of Observations:                              15
Sample Minimum:                                  -1.400
ID for Sample Minimum:                                1
Sample Maximum:                                   1.010
ID for Sample Maximum:                               15
Sample Mean:                                      0.018
Sample SD:                                        0.551
Sample Range:                                     2.410
 
David Test Statistic Value:                       4.374
 
 
Conclusions (Upper 1-Tailed Test)
-------------------------------------------------------------
  Alpha    CDF      Statistic   Critical Value     Conclusion
-------------------------------------------------------------
    10%    90%          4.374            4.025      Reject H0
     5%    95%          4.374            4.171      Reject H0
     1%    99%          4.374            4.435      Accept H0
 
 
 
Critical Values Based on ASTM E-178 Tables
 
 
THE FORTRAN COMMON CHARACTER VARIABLE DAVITEST HAS JUST BEEN SET TO DAVI
 
            David, Hartley, Pearson Test for Outliers: Simultaneous Test
                  for Minimum and Maximum (Assumption: Normality)
 
Response Variable: Y
 
H0: The minimum and maximum are not
    both outliers
Ha: Both the minimum and maximum are
    outliers
Potential minimum outlier value tested:          -1.400
Potential maximum outlier value tested:           1.010
 
Summary Statistics:
Number of Observations:                              15
Sample Minimum:                                  -1.400
ID for Sample Minimum:                                1
Sample Maximum:                                   1.010
ID for Sample Maximum:                               15
Sample Mean:                                      0.018
Sample SD:                                        0.551
Sample Range:                                     2.410
 
David Test Statistic Value:                       4.374
 
 
Conclusions (Upper 1-Tailed Test)
-------------------------------------------------------------
  Alpha    CDF      Statistic   Critical Value     Conclusion
-------------------------------------------------------------
    10%    90%          4.374            4.020      Reject H0
     5%    95%          4.374            4.170      Reject H0
   2.5%  97.5%          4.374            4.290      Reject H0
     1%    99%          4.374            4.430      Accept H0
   0.5%  99.5%          4.374            4.530      Accept H0
 
 
 
Critical Values Based on David Tables
 
 
THE FORTRAN COMMON CHARACTER VARIABLE DAVITEST HAS JUST BEEN SET TO FORM
 
            David, Hartley, Pearson Test for Outliers: Simultaneous Test
                  for Minimum and Maximum (Assumption: Normality)
 
Response Variable: Y
 
H0: The minimum and maximum are not
    both outliers
Ha: Both the minimum and maximum are
    outliers
Potential minimum outlier value tested:          -1.400
Potential maximum outlier value tested:           1.010
 
Summary Statistics:
Number of Observations:                              15
Sample Minimum:                                  -1.400
ID for Sample Minimum:                                1
Sample Maximum:                                   1.010
ID for Sample Maximum:                               15
Sample Mean:                                      0.018
Sample SD:                                        0.551
Sample Range:                                     2.410
 
David Test Statistic Value:                       4.374
 
 
Conclusions (Upper 1-Tailed Test)
-------------------------------------------------------------
  Alpha    CDF      Statistic   Critical Value     Conclusion
-------------------------------------------------------------
    20%    80%          4.374            3.875      Reject H0
    10%    90%          4.374            4.034      Reject H0
     5%    95%          4.374            4.173      Reject H0
   2.5%  97.5%          4.374            4.295      Reject H0
     1%    99%          4.374            4.435      Accept H0
   0.5%  99.5%          4.374            4.527      Accept H0
 
 
 
Critical Values Based on Formula
 
 
THE FORTRAN COMMON CHARACTER VARIABLE DAVITEST HAS JUST BEEN SET TO SIMU
 
            David, Hartley, Pearson Test for Outliers: Simultaneous Test
                  for Minimum and Maximum (Assumption: Normality)
 
Response Variable: Y
 
H0: The minimum and maximum are not
    both outliers
Ha: Both the minimum and maximum are
    outliers
Potential minimum outlier value tested:          -1.400
Potential maximum outlier value tested:           1.010
 
Summary Statistics:
Number of Observations:                              15
Sample Minimum:                                  -1.400
ID for Sample Minimum:                                1
Sample Maximum:                                   1.010
ID for Sample Maximum:                               15
Sample Mean:                                      0.018
Sample SD:                                        0.551
Sample Range:                                     2.410
 
David Test Statistic Value:                       4.374
CDF Value:                                        0.986
P-Value                                           0.014
 
 
 
Conclusions (Upper 1-Tailed Test)
-------------------------------------------------------------
  Alpha    CDF      Statistic   Critical Value     Conclusion
-------------------------------------------------------------
    20%    80%          4.374            3.842      Reject H0
    10%    90%          4.374            4.021      Reject H0
     5%    95%          4.374            4.166      Reject H0
   2.5%  97.5%          4.374            4.296      Reject H0
     1%    99%          4.374            4.428      Accept H0
   0.5%  99.5%          4.374            4.523      Accept H0
 
 
 
Critical Values Based on 50,000 Simulations