Hot to set a constraint on the guessing parameter

[Balazs Klein]

Hi,

how can I set the constraint of the guessing parameter of a 3PL model to be the same for all items?

I only know now how to set the guessing parameter of all items to be some value.

fit <- mirt(select_data, 1, itemtype ='3PL', guess = 0.1)

Now I would like the algorithm to come up with that value instead.

Thanks.

Balázs

[Phil Chalmers]

Hello!

I think I am doing something wrong here. But how do I constraint the guessing to be the same for all items? 

library(mirt)

dat <- expand.table(LSAT7)

mod <- mirt(dat, 1,itemtype ='3PL', guess = 0.3, verbose=FALSE)

coef(mod, IRTpar=T, simplify=TRUE)

$items

                 a      b          g   u

Item.1 1.007 -1.852 0.001 1

Item.2 1.927 -0.049 0.295 1

Item.3 1.669 -1.066 0.002 1

Item.4 0.736 -0.654 0.000 1

Item.5 0.769 -2.419 0.008 1

Thanks,

Aiden