TY - JOUR

T1 - Some new results on correlation-preserving factor scores prediction methods

AU - ten Berge, Jos M.F.

AU - Krijnen, Wim P.

AU - Wansbeek, Tom

AU - Shapiro, Alexander

PY - 1999/3/1

Y1 - 1999/3/1

N2 - Anderson and Rubin and McDonald have proposed a correlation-preserving method of factor scores prediction which minimizes the trace of a residual covariance matrix for variables. Green has proposed a correlation-preserving method which minimizes the trace of a residual covariance matrix for factors. Krijnen, Wansbeek and Ten Berge have proposed minimizing the determinant rather than the trace of the latter covariance matrix, and offered an iterative procedure to that effect. In the present paper it is shown that the iterative procedure can be replaced by a closed-form solution. When all unique variances are strictly positive, this solution is the same as McDonald's. The solution coincides with Green's solution in certain special cases, for instance, when the factors are orthogonal.

AB - Anderson and Rubin and McDonald have proposed a correlation-preserving method of factor scores prediction which minimizes the trace of a residual covariance matrix for variables. Green has proposed a correlation-preserving method which minimizes the trace of a residual covariance matrix for factors. Krijnen, Wansbeek and Ten Berge have proposed minimizing the determinant rather than the trace of the latter covariance matrix, and offered an iterative procedure to that effect. In the present paper it is shown that the iterative procedure can be replaced by a closed-form solution. When all unique variances are strictly positive, this solution is the same as McDonald's. The solution coincides with Green's solution in certain special cases, for instance, when the factors are orthogonal.

U2 - 10.1016/s0024-3795(97)10007-6

DO - 10.1016/s0024-3795(97)10007-6

M3 - Article

VL - 289

SP - 311

EP - 318

JO - Linear Algebra and Its Applications

JF - Linear Algebra and Its Applications

SN - 0024-3795

IS - 1-3

ER -