Analyzes a completely nested random model with possibly unequal numbers in the subgroups.
NF — Number of
factors (number of subscripts) in the model including
error.
(Input)
IEQ — Equal numbers option. (Input)
IEQ Description
0 Unequal numbers in the subgroups
1 Equal numbers in the subgroups
NL — Vector with the number of
levels. (Input)
If IEQ = 1, NL
is of length NF
and contains the number of levels for each of the factors. In this case, the
following additional variables are referred to in the description of ANEST:
Variable Description
LNL NL(1) + NL(1) * NL(2) + … + NL(1) * NL(2) * … * NL(NF − 1)
LNLNF NL(1) * NL(2) * …* NL(NF − 1)
NOBS
The number of observations. NOBS equals NL(1)
* NL(2)
* …
* NL(NF).
If IEQ = 0, NL contains the number of levels of each factor at each level of the factor in which it is nested. In this case, the following additional variables are referred to in the description of ANEST:
Variable Description
LNL Length of NL.
LNLNF Length of the subvector of NL for the last factor.
NOBS
Number of observations. NOBS
equals the sum of the last
LNLNF elements of NL.
For example, a random one-way model with two groups, five responses in the first
group and ten in the second group, would have LNL = 3, LNLNF = 2, NOBS = 15,
NL(1)
= 2, NL(2)
= 5, and NL(3)
= 10.
Y — Vector of
length NOBS
containing the responses. (Input)
The elements of Y
are ordered lexicographically, i.e., the last model subscript changes most
rapidly, the next next to last model subscript changes the next most rapidly,
and so forth, with the first subscript changing the slowest.
AOV — Vector of length 15 containing statistics relating to the analysis of variance. (Output)
I |
AOV(I) |
1 |
Degrees of freedom for regression |
2 |
Degrees of freedom for error |
3 |
Total degrees of freedom |
4 |
Sum of squares for regression |
5 |
Sum of squares for error |
6 |
Total sum of squares |
7 |
Regression mean square |
8 |
Error mean square |
9 |
F-statistic |
10 |
p-value |
11 |
R2 (in percent) |
12 |
Adjusted R2 (in percent) |
13 |
Estimated standard deviation of the model error |
14 |
Mean of the response (dependent) variable |
15 |
Coefficient of variation (in percent) |
CONPER —
Confidence level for two-sided interval estimates on the variance components, in
percent. (Input)
Default: CONPER
= 95.0.
CONPER percent
confidence intervals are computed, hence, CONPER must be in the
interval [0.0, 100.0). CONPER often will be
90.0, 95.0, or 99.0. For one-sided intervals with confidence level ONECL, ONECL in the interval
[50.0, 100.0), set
CONPER = 100.0 − 2.0 * (100.0 − ONECL).
IPRINT — Printing
option. (Input)
Default: IPRINT
= 0.
IPRINT |
Action |
0 |
No printing is performed. |
1 |
Printing is performed. |
EMS — Vector of length (NF + 1) * NF/2 with expected mean square coefficients. (Output)
VC — NF
by 9 matrix containing statistics relating to the particular variance components
in the model. (Output)
Rows of VC correspond to the
NF
factors. Columns of VC are as follows:
Column |
Description |
1 |
Degrees of freedom |
2 |
Sum of squares |
3 |
Mean squares |
4 |
F-statistic |
5 |
p-value for F test |
6 |
Variance component estimate |
7 |
Percent of variance explained by variance component |
8 |
Lower endpoint for a confidence interval on the variance component |
9 |
Upper endpoint for a confidence interval on the variance component |
A test for the error variance
equal to zero cannot be performed. VC(NF, 4) and
VC(NF, 5) are set to NaN
(not a number).
LDVC — Leading
dimension of VC
exactly as specified in the dimension statement in the calling
program. (Input)
Default: LDVC= size(VC ,1)
YMEANS — Vector containing the subgroup means. (Output)
IEQ Length of YMEANS
0 1 + NL(1) + NL(2) + … NL(LNL − LNLNF) (See the description of argument NL for definitions of LNL and LNLNF.)
1 1 + NL(1) + NL(1) * NL(2) + … + NL(1) * NL(2) * … * NL(NF − 1)
If the factors are labeled A, B, C, and error, the ordering
of the means is grand mean,
A means, AB means, and then
ABC means.
NMISS — Number of
missing values in Y.
(Output)
Elements of Y
equal to NaN (not a number) are omitted from the computations.
Generic: CALL ANEST (NF, IEQ, NL, Y, AOV [,…])
Specific: The specific interface names are S_ANEST and D_ANEST.
Single: CALL ANEST (NF, IEQ, NL, Y, CONPER, IPRINT, AOV, EMS, VC, LDVC, YMEANS, NMISS)
Double: The double precision name is DANEST.
Routine ANEST analyzes a nested random model with equal or unequal numbers in the subgroups. The analysis includes an analysis of variance table and computation of subgroup means and variance component estimates. Anderson and Bancroft (1952, pages 325−330) discuss the methodology. The analysis of variance method is used for estimating the variance components. This method solves a linear system in which the mean squares are set to the expected mean squares. A problem that Hocking (1985, pages 324−330) discusses is that this method can yield negative variance component estimates. Hocking suggests a diagnostic procedure for locating the cause of a negative estimate. It may be necessary to reexamine the assumptions of the model.
Workspace may be explicitly provided, if desired, by use of A2EST/DA2EST. The reference is:
CALL A2EST (NF, IEQ, NL, Y, CONPER, IPRINT, AOV, EMS, VC, LDVC, YMEANS, NMISS, WK, IWK, CHWK)
The additional arguments are as follows:
WK — Work vector of length NOBS.
IWK — Work vector of length 5 * NF + (2 * LNL − LNLNF).
CHWK — CHARACTER * 10 vector of length 2 * NF + 1. If IPRINT = 0, CHWK is not referenced and can be a vector of length one.
An analysis of a three-factor nested random model with equal numbers in the subgroups is performed using data discussed by Snedecor and Cochran (1967, Table 10.16.1, pages 285−288). The responses are calcium concentrations (in percent, dry basis) as measured in the leaves of turnip greens. Four plants are taken at random, then three leaves are randomly selected from each plant. Finally, from each selected leaf two samples are taken to determine calcium concentration. The model is
yijk = μ + α i + βjj + eijk i = 1, 2, 3, 4; j = 1, 2, 3; k = 1, 2
where yijk is the calcium concentration for the k-th sample of the j-th leaf of the i-th plant, the α i’s are the plant effects and are taken to be independently distributed
the βij’s are leaf effects each independently distributed
and the ɛ ijk’s are errors each independently distributed N(0, σ2). The effects are all assumed to be independently distributed. The data are given in the following table:
Plant |
Leaf |
Samples | |
1 |
1 2 3 |
3.28 3.52 2.88 |
3.09 3.48 2.80 |
2 |
1 2 3 |
2.46 1.87 2.19 |
2.44 1.92 2.19 |
3 |
1 2 3 |
2.77 3.74 2.55 |
2.66 3.44 2.55 |
4 |
1 2 3 |
3.78 4.07 3.31 |
3.87 4.12 3.31 |
USE ANEST_INT
IMPLICIT NONE
INTEGER NF, NOBS
PARAMETER (NF=3, NOBS=24)
!
INTEGER IEQ, IPRINT, NL(NF)
REAL AOV(15), Y(NOBS)
!
DATA Y/3.28, 3.09, 3.52, 3.48, 2.88, 2.80, 2.46, 2.44, 1.87, &
1.92, 2.19, 2.19, 2.77, 2.66, 3.74, 3.44, 2.55, 2.55, 3.78, &
3.87, 4.07, 4.12, 3.31, 3.31/
DATA NL/4, 3, 2/
!
IEQ = 1
IPRINT = 1
CALL ANEST (NF, IEQ, NL, Y, AOV, IPRINT=IPRINT)
END
Dependent R-squared Adjusted Est. Std.
Dev.
Coefficient of
Variable (percent) R-squared of Model
Error Mean Var.
(percent)
Y
99.222
98.510
0.08158
3.012 2.708
* * * Analysis of Variance * * *
Sum of Mean Prob. of
Source DF Squares Square Overall F Larger F
Model 11 10.19 0.9264 139.216 0.0000
Error 12 0.08 0.0067
Corrected Total 23 10.27
Sum of
Mean
Prob. of
Source DF Squares Square F Larger F
A 3 7.56034 2.52011 7.665 0.0097
B 8 2.63020 0.32878 49.406 0.0000
* * * Expected Mean Square Coefficients * * *
Error Effect B Effect A
Effect A 1 2 6
Effect B 1 2
Error 1
* * * Variance Components * * *
95.0% Confidence Interval
Variance --------------------------
Component Estimate Percent Lower Limit Upper Limit
A 0.36522 68.530 0.039551 5.7867
B 0.16106 30.221 0.069669 0.6004
Error 0.00665 1.249 0.003422 0.0181
A Means
1 3.1750
2 2.1783
3 2.9517
4 3.7433
AB Means
1 1 3.1850
1 2 3.5000
1 3 2.8400
2 1 2.4500
2 2 1.8950
2 3 2.1900
3 1 2.7150
3 2 3.5900
3 3 2.5500
4 1 3.8250
4 2
4.0950
4 3 3.3100
An analysis of a three-factor nested random model with unequal numbers in the subgroups is performed. The data are given in the following table:
A |
B |
C | ||
1 |
1 2 |
23.0 31.0 |
19.0 37.0 |
|
2 |
1 2 |
33.0 29.0 |
29.0 |
|
3 |
1 |
36.0 |
29.0 |
33.0 |
4 |
1 2 3 4 5 6 7 8 9 |
11.0 23.0 33.0 23.0 26.0 39.0 20.0 24.0 36.0 |
21.0 18.0 |
|
5 |
1 |
25.0 |
33.0 |
|
6 |
1 2 3 4 5 6 7 8 9 10 |
28.0 25.0 32.0 41.0 35.0 16.0 30.0 40.0 32.0 44.0 |
31.0 42.0 36.0 |
|
USE ANEST_INT
IMPLICIT NONE
INTEGER LNL, NF, NOBS
PARAMETER (LNL=32, NF=3, NOBS=36)
!
INTEGER IEQ, IPRINT, NL(LNL)
REAL AOV(15), Y(NOBS)
!
DATA Y/23.0, 19.0, 31.0, 37.0, 33.0, 29.0, 29.0, 36.0, 29.0, &
33.0, 11.0, 21.0, 23.0, 18.0, 33.0, 23.0, 26.0, 39.0, 20.0, &
24.0, 36.0, 25.0, 33.0, 28.0, 31.0, 25.0, 42.0, 32.0, 36.0, &
41.0, 35.0, 16.0, 30.0, 40.0, 32.0, 44.0/
DATA NL/6, 2, 2, 1, 9, 1, 10, 2, 2, 2, 1, 3, 2, 2, 1, 1, 1, 1, &
1, 1, 1, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1/
!
IEQ = 0
IPRINT = 1
CALL ANEST (NF, IEQ, NL, Y, AOV, IPRINT=IPRINT)
END
Dependent R-squared Adjusted Est. Std.
Dev.
Coefficient of
Variable (percent) R-squared of Model
Error Mean Var.
(percent)
Y
85.376
53.470
5.31
29.53 17.98
* * * Analysis of Variance * * *
Sum of Mean Prob. of
Source DF Squares Square Overall F Larger F
Model 24 1810.8 75.45 2.676 0.0459
Error 11 310.2 28.20
Corrected Total 35 2121.0
Sum
of
Mean
Prob. of
Source DF Squares Square F Larger F
A 5 461.42 92.2845 0.988 0.4601
B 19 1349.38 71.0202 2.519 0.0597
* * * Expected Mean Square Coefficients * * *
Error Effect B Effect A
Effect A 1.00000 1.96503 5.37778
Effect B 1.00000 1.28990
Error 1.00000
* * * Variance Components * * *
95.0% Confidence Interval
Variance --------------------------
Component Estimate Percent Lower Limit Upper Limit
A -0.214 NaN NaN NaN
B 33.199 54.073 0.00 100.59
Error 28.197 45.927 14.15 81.29
A Means
1 27.5000
2 30.3333
3 32.6667
4 24.9091
5 29.0000
6
33.2308
AB Means
1 1 21.0000
1 2 34.0000
2 1 31.0000
2 2 29.0000
3 1 32.6667
4 1 16.0000
4 2 20.5000
4 3 33.0000
4 4 23.0000
4 5 26.0000
4 6 39.0000
4 7 20.0000
4 8 24.0000
4 9 36.0000
5 1 29.0000
6 1 29.5000
6 2 33.5000
6 3 34.0000
6 4 41.0000
6 5 35.0000
6 6 16.0000
6 7 30.0000
6 8 40.0000
6 9 32.0000
6 10 44.0000
PHONE: 713.784.3131 FAX:713.781.9260 |