generate_item_responses_sml

theta

matrix describing scoring patterns on each latent trait

scoringMatrix

vector of slope parameters of each trait

slopes

intercepts

Function generates responses 'simultaneously` with a whole
scoring matrix used at once. Only (G)PCM approach is suitable in such a case,
because with complicated scoring matrices there is no guarantee that
probabilities of responses are increasing along with order of responses
(rows) in a scoring matrix. Consequently, no normal ogive models can be used.

Package allows to generate simulated datasets using algorithms
that mimic different response styles to survey questions using: 1) IRTree
approach (Bockenholt (2012) <doi:10.1037/a0028111>,
(2017) <doi:10.1037/met0000106>), 2) (G)PCM (and rating scale version of
a partial credit model) random-thresholds approach (Falk & Cai (2016)
<doi:10.1037/met0000059>; Henninger & Meiser (2020a)
<doi:10.1037/met0000249>, (2020b) <doi:10.1037/met0000268>; Plieninger
(2017) <doi:10.1177/0013164416636655>), 3) user provided function that (with
some probability) chooses response using information about previous
responses. This allows to cover wide range of potential response styles
like: extreme and middle (ERS, MRS), acquiesce (ARS) and also
careless/inattentive responding (CR, IR).

generate_item_responses_sml: Internals: simulating responding to item in a 'simultaneous' way

Description

Usage

Arguments

Value

See Also