Compute E-step for Binary Mediator Misclassification Model Estimated With the EM Algorithm

Note that this function should only be used for Normal outcome models.

Usage

w_m_normalY(
  mstar_matrix,
  pistar_matrix,
  pi_matrix,
  p_yi_m0,
  p_yi_m1,
  sample_size,
  n_cat
)

Arguments

mstar_matrix: A numeric matrix of indicator variables (0, 1) for the observed mediator M*. Rows of the matrix correspond to each subject. Columns of the matrix correspond to each observed mediator category. Each row should contain exactly one 0 entry and exactly one 1 entry.
pistar_matrix: A numeric matrix of conditional probabilities obtained from the internal function pistar_compute. Rows of the matrix correspond to each subject and to each observed mediator category. Columns of the matrix correspond to each true, latent mediator category.
pi_matrix: A numeric matrix of probabilities obtained from the internal function pi_compute. Rows of the matrix correspond to each subject. Columns of the matrix correspond to each true, latent mediator category.
p_yi_m0: A numeric vector of Normal outcome likelihoods computed assuming a true mediator value of 0.
p_yi_m1: A numeric vector of Normal outcome likelihoods computed assuming a true mediator value of 1.
sample_size: An integer value specifying the number of observations in the sample. This value should be equal to the number of rows of the observed mediator matrix, mstar_matrix.
n_cat: The number of categorical values that the true outcome, M, and the observed outcome, M*, can take.

Value

w_m_normalY returns a matrix of E-step weights for the EM-algorithm. Rows of the matrix correspond to each subject. Columns of the matrix correspond to the true mediator categories \(j = 1, \dots,\) n_cat.