simulate_morbist: Simulate test responses in the MORBIST model
In m-Py/DOTA2: Discrete option test analysis

Description Usage Arguments Value Author(s) Examples

Simulates a specified number of test-takers working on a specified number of items in MC or DOMC format. The accuracy values d' of test-takers can be generated randomly by the function or are passed explicitly.

1
2
3

simulate_morbist(item_number, option_number, n_respondents, type,
  avg_acc = 1.5, sd_acc = 0.5, accuracies = NULL, criterion_list = NULL,
  option_data = FALSE)

`item_number`	How many items are simulated for each test-taker
`option_number`	How many response options should each item have
`n_respondents`	How many test-takers are to be simulated
`type`	Which type of test is be simulated? Must be "mc" or "domc".
`avg_acc`	Mean accuracy (d') of the test-takers
`sd_acc`	Standard deviation of the test-takers' accuracy values d'.
`accuracies`	If not NULL, it must be a vector of length ‘n_respondents' containing an accuracy value d’ for each test taker. If this argument is passed, it overrides the values of avg_acc and sd_acc. Defaults to NULL.
`criterion_list`	A list of criterion vectors of length 'testTaker'.
`option_data`	Should response data on option level be returned (only works if type is "domc"). This is slows down simulation speed; it defaults to 'FALSE'.

If a DOMC test was simulated and option level data was requested: A list containing two data frames in long format - one stores data on item level, the other store response data on option level. Otherwise: A data.frame in long format storing response data on item level.

Martin Papenberg martin.papenberg@hhu.de

## Prepare input
n = 100
n_items <- 50
n_options <- 5
criteria <- list() # response criteria for DOMC items
for (i in 1:n)  criteria[[i]] <- c(rep(0, 4), -Inf)

## Simulate DOMC test responses
DOMC_data <- simulate_morbist(n_items, n_options, n, type="domc",
                              criterion_list = criteria)

## Simulate MC test responses
MC_data <- simulate_morbist(n_items, n_options, n, type="mc")