search

drizopoulos / ltm

Latent Trait Models under IRT
30 stars 12 forks source link

the Arguments "return.MIvalues" in the function factor.scores is no use, when I let it TRUE or FALSE, it's result is same #10

Open ICESDHR opened 5 years ago

ICESDHR commented 5 years ago

Don't have extra attributes "zvalues.MI" and "var.zvalues.MI"

PaulTMcG commented 5 years ago

I have the same problem. Here's a reproduce-able example: set.seed(382) df <- bind_rows(data.frame(one = sample(c(4, 5), 100, replace = T), two = sample(c(4, 5), 100, replace = T), three = sample(c(4, 5), 100, replace = T), four = sample(c(4, 5), 100, replace = T), five = sample(c(4, 5), 100, replace = T)), data.frame(one = sample(c(3, 4, 5), 150, replace = T), two = sample(c(3, 4, 5), 150, replace = T), three = sample(c(3, 4, 5), 150, replace = T), four = sample(c(3, 4, 5), 150, replace = T), five = sample(c(3, 4, 5), 150, replace = T)), data.frame(one = sample(c(2, 3, 4), 250, replace = T), two = sample(c(2, 3, 4), 250, replace = T), three = sample(c(2, 3, 4), 250, replace = T), four = sample(c(2, 3, 4), 250, replace = T), five = sample(c(2, 3, 4), 250, replace = T)), data.frame(one = sample(c(1, 2, 3), 250, replace = T), two = sample(c(1, 2, 3), 250, replace = T), three = sample(c(1, 2, 3), 250, replace = T), four = sample(c(1, 2, 3), 250, replace = T), five = sample(c(1, 2, 3), 250, replace = T)), data.frame(one = sample(c(0, 1, 2), 150, replace = T), two = sample(c(0, 1, 2), 150, replace = T), three = sample(c(0, 1, 2), 150, replace = T), four = sample(c(0, 1, 2), 150, replace = T), five = sample(c(0, 1, 2), 150, replace = T)), data.frame(one = sample(c(0, 1), 100, replace = T), two = sample(c(0, 1), 100, replace = T), three = sample(c(0, 1), 100, replace = T), four = sample(c(0, 1), 100, replace = T), five = sample(c(0, 1), 100, replace = T)) )

fit_df <- grm(df, Hessian = TRUE) scores_df <- factor.scores.grm(fit_df) scores_mi_df <- factor.scores.grm(fit_df, method = "MI", B = 10, return.MIvalues = TRUE)