Login

News

Confidence intervals do not agree with estimates! Should not happen

Submitted by neale on

free parameters:
name matrix row col Estimate Std.Error lbound ubound
1 b11 MZ.b 1 1 2.7536823 0.69611278
2 m1MZ MZ.meanMZ 1 1 20.6886950 0.16960595
3 m2MZ MZ.meanMZ 1 2 20.6934722 0.16988419
4 v1MZ MZ.expCovMZ bmi1 bmi1 0.7213571 0.04326283 1e-04
5 c21MZ MZ.expCovMZ bmi1 bmi2 0.5840681 0.04034537 0
6 v2MZ MZ.expCovMZ bmi2 bmi2 0.7842634 0.04713233 1e-04
7 m1DZ DZ.meanDZ 1 1 20.7833538 0.17188984
8 m2DZ DZ.meanDZ 1 2 20.8080580 0.17250024
9 v1DZ DZ.expCovDZ bmi1 bmi1 0.7281360 0.05594797 1e-04
10 c21DZ DZ.expCovDZ bmi1 bmi2 0.2414624 0.04378431 0
11 v2DZ DZ.expCovDZ bmi2 bmi2 0.8030416 0.06183095 1e-04

confidence intervals:
lbound estimate ubound
MZ.expCovMZ[1,1] 0.6427435 0.7213571 0.8133005
MZ.expCovMZ[1,2] 0.5105505 0.5840681 0.6696089
MZ.expCovMZ[2,1] 0.5105505 0.5840681 0.6696089
MZ.expCovMZ[2,2] 0.6986221 0.7842634 0.8844486
DZ.expCovDZ[1,1] 0.6285763 0.7281360 0.8499117
DZ.expCovDZ[1,2] 0.1596200 0.2414624 0.3333679
DZ.expCovDZ[2,1] 0.1596200 0.2414624 0.3333679
DZ.expCovDZ[2,2] 0.6930290 0.8030416 0.9376731
MZ.expMeanMZ[1,1] 21.4045930 21.1568210 21.6250361
MZ.expMeanMZ[1,2] 21.4073401 21.1615982 21.6318426
DZ.expMeanDZ[1,1] 21.4846632 21.2514798 21.7341197
DZ.expMeanDZ[1,2] 21.5058433 21.2761840 21.7623907

Why are the means different in the two sections?

The confidence intervals agree fine. There's a matrix meanMZ and an algebra expMeanMz.


twinSatFit$submodels$MZ$matrices$meanMZ$values
# [,1] [,2]
#[1,] 20.6887 20.69348
twinSatFit$submodels$MZ$algebras$expMeanMZ$result
# [,1] [,2]
#[1,] 21.5148 21.51958
twinSatFit$submodels$MZ$algebras$expMeanMZ
#mxAlgebra 'expMeanMZ'
#$formula: meanMZ + cbind(b %*% Age, b %*% Age)
#$result:
# [,1] [,2]
#[1,] 21.5148 21.51958
#dimnames: NULL

The parameters in summary are from the matrix, but the confidence interval is on the algebra.

> mxEval(MZ.meanMZ, twinSatFit, compute=TRUE)
[,1] [,2]
[1,] 20.6887 20.69348
> mxEval(MZ.expMeanMZ, twinSatFit, compute=TRUE)
[,1] [,2]
[1,] 21.15682 21.1616

Log in or register to post comments
  Login

News