maat is the main function for simulating a multi-stage multi-administration adaptive test.
maat(
examinee_list = examinee_list,
assessment_structure = NULL,
module_list = NULL,
config = NULL,
cut_scores = NULL,
overlap_control_policy = NULL,
transition_policy = "CI",
combine_policy = "conditional",
transition_CI_alpha = NULL,
transition_percentile_lower = NULL,
transition_percentile_upper = NULL,
initial_theta_list = NULL,
prior_mean_policy = "mean_difficulty",
prior_mean_user = NULL,
prior_sd = 1,
verbose = TRUE
)an examinee list from simExaminees.
a assessment_structure object.
a module list from loadModules.
a config_Shadow object. Also accepts a list of config_Shadow objects to use separate configurations for each module. Must be from 'TestDesign' 1.3.3 or newer, and its exclude_policy$method slot must be SOFT.
a named list containing cut scores to be used in each grade. Each element must be named in the form G?, where ? is a number.
overlap control is performed by excluding administered items from being administered again within the same examinee.
all performs overlap control at all module positions.
within_test performs overlap control only within each test.
none does not perform overlap control.
CI uses the confidence interval to perform routing.
pool_difficulty_percentile uses item difficulty percentiles of all items in the item_pool argument to perform routing.
pool_difficulty_percentile_exclude_administered uses item difficulty percentiles of all items in the item_pool argument to perform routing, excluding all previous items administered to the examinee.
on_grade does not permit any transition.
(default = CI)
This is only applied when module_position %% 2 == 0 (at Phase 2, which is the end of each test).
conditional uses the combined theta (using items from the previous module combined with the current module), if the examinee was in the same grade in Phases 1 and 2. If the examinee was in different grades in Phases 1 and 2, then the theta estimate from Phase 2 is used.
always uses the combined theta.
never uses the theta estimate from Phase 2.
(default = conditional)
the alpha level to use when transition_policy == "CI".
the percentile value (between 0 and 1) to use for the lower routing when transition_policy == "difficulty_percentile".
the percentile value (between 0 and 1) to use for the upper routing when transition_policy == "difficulty_percentile".
(optional) a list containing initial thetas to use in each module position.
This is only effective at the beginning of each test. This determines what value is used as the prior mean.
mean_difficulty uses the mean item difficulty of the current item pool.
carryover uses the routing theta from the previous module. For Phase 1 of the first test, user supplied values are used if available. Otherwise, the mean item difficulty of the current item pool is used.
user uses user-supplied values in the prior_mean_user argument.
(default = mean_difficulty)
(optional) user-supplied values for the prior mean. Must be a single value, or a vector for each grade.
user-supplied values for the prior standard deviation. This is only effective at the beginning of each test. This is utilized regardless of prior_mean_policy. Must be a single value, or a vector for each grade. (default = 1)
if TRUE, print status messages. (default = TRUE)
an output_maat object from the simulation.
# \donttest{
library(TestDesign) # >= 1.3.3
config <- createShadowTestConfig(
final_theta = list(method = "MLE"),
exclude_policy = list(method = "SOFT", M = 100)
)
# exclude_policy must be SOFT
examinee_list <- maat(
examinee_list = examinee_list_math,
assessment_structure = assessment_structure_math,
module_list = module_list_math,
overlap_control_policy = "all",
transition_CI_alpha = 0.05,
config = config,
cut_scores = cut_scores_math
)
# }