Compute process variance but with minor error in Jaech Fortran code.

Description

This is an internal function to compute the process variance that replicates the minor error in Jaech's Fortran code. This allows comparing merror estimates to those shown in Jaech 1985.

Usage

1
process.var.mle.jaech.err(sigma2, s, beta.bars, N, n)

Arguments

sigma2

Estimated imprecisions for each method in an N vector.

s

Variance-covariance N x N matrix.

beta.bars

Estimates or hypothesized values for the N betas

N

No. of methods.

n

No. of items.

Details

See Jaech p. 186 equations 6.37 - 6.3.10. Jaech p. 288 line 2330 has s[i,j] instead of s[j,j]. Jaech p. 288 line 2410 omits "- 1/d2".

Value

Estimated process variance but replicating minor error in Jaech's Fortran code.

Author(s)

Richard A. Bilonick

References

Jaech, J. L. (1985) Statistical Analysis of Measurement Errors. New York: Wiley.

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.