Skip to contents

Returns the assigned booklets and units for testtakers in a Testcenter workspace. Supplying units expands the design to variables and restricts it to units with available variables. Supply all units used in the test. With units = NULL, the result remains at unit level, without variable_id.

Usage

get_design(workspace, units = NULL, overwrite = FALSE, mode = "run-hot-return")

# S4 method for class 'WorkspaceTestcenter'
get_design(workspace, units = NULL, overwrite = FALSE, mode = "run-hot-return")

Arguments

workspace

WorkspaceTestcenter. Workspace information necessary to retrieve design information.

units

Tibble (optional). Unit data retrieved from get_units(). Coding schemes supply the variables for each unit. Units without a coding scheme use unit_variables$variable_id instead. If these are unavailable, explicit variable IDs in a StarS unit_definition are used. Retrieve definitions with get_units(..., unit_definition = TRUE). May also contain unit_codes prepared by add_coding_scheme() to save time.

overwrite

Logical. Should existing unit_codes be rebuilt? Defaults to FALSE, which reuses codes prepared by add_coding_scheme().

mode

Character. Only testtakers with the specified modes will be filtered. Defaults to "run-hot-return".

Value

A tibble.

Details

Variable selection is performed separately for each unit, supporting mixed tests with and without coding schemes. Existing schemes supply their base and derived variables, excluding deactivated (BASE_NO_VALUE) variables. Only a missing or empty scheme (without variableCodings) allows a fallback to unit_variables; a scheme containing only deactivated variables does not. Cached unit_codes are reused unless overwrite = TRUE. Empty cached results allow the fallback only if the underlying scheme is missing or empty; otherwise they remain empty until rebuilt with overwrite = TRUE. The fallback uses the normalized IDs from get_units(), including its alias handling. Duplicate IDs are collapsed and missing or blank IDs are ignored. If no usable metadata IDs remain, StarS definitions supply explicit IDs from interactionParameters$variableId, variableInfo, and the selection/outcome fields of closingMetaButtons. Reference IDs and unrelated JSON fields are not included. Implicit player defaults are not inferred; definitions must explicitly name the variables needed for the design. Units without any available variable IDs contribute no variable-level rows.

The variable list comes from metadata or definitions, not observed responses, so it can include variables that nobody answered. For StarS data, check that the IDs in the metadata or definitions match those returned by prepare_coded(). A variable-level design alone does not supply the coding metadata required by complete_design().

Functions

  • get_design(WorkspaceTestcenter): Get design in a Testcenter workspace