obtain_nrm_def | R Documentation |
mirt
's nominal model taking in account the key
of correct answersStandard mirt
model with itemtype = "nominal"
puts the
identification constrains on the item response category slopes such as
ak_0 = 0
and ak_{(K-1)} = (K - 1)
, freely estimating the rest.
While nominal item responses are unordered by definition, it is often the
case that one of the item response categories is correct and the
respondents endorsing this category "naturally" possess a higher latent
ability. Use this function to obtain model definition where the correct
response category k_c
for item i
with K
possible response
categories translates to constrains ak_{k_c} = (K - 1)
and
ak_{k_{d1}} = 0
, with k_{d1}
being the first incorrect response
category (i.e. the first distractor).
obtain_nrm_def(data_with_key, ...)
data_with_key |
The output of |
... |
arguments passed onto |
A data.frame
with the starting values, parameter numbers,
estimation constrains etc. Pass it as pars
argument of mirt::mirt()
.
Other BLIS/BLIRT related:
BlisClass-class
,
coef,BlisClass-method
,
fit_blis()
,
get_orig_levels()
,
nominal_to_int()
,
print.blis_coefs()
library(mirt)
# convert nominal data to integers and the original labels with correct answers
data_with_key <- nominal_to_int(HCItest[, 1:20], HCIkey)
# build model definition for {mirt} using the returned list from above
nrm_def <- obtain_nrm_def(data_with_key)
# fit the nominal model using the obtained model definition in `pars` argument
fit <- mirt(data_with_key$Data, 1, "nominal", pars = nrm_def)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.