Skip to contents

Create a Bring-Your-Own O*NET Measure

Source: R/measure.R
onet_measure.Rd

Validates a user-supplied occupation, task, or DWA score table and records coverage against an optional universe. The package does not supply or alter the substantive score.

Usage

onet_measure(
  data,
  key = NULL,
  score = NULL,
  key_type = c("occupation", "task", "dwa"),
  universe = NULL,
  measure_id = "user_measure",
  measure_name = measure_id,
  source = NA_character_,
  release_version = NA_character_,
  weight_panel = NULL,
  items = NULL,
  agg = NULL,
  item = "task_id",
  scale = "IM"
)

Arguments

data

A data frame. For the default path, the user-supplied measure table. For the items/agg path, a Task Ratings style panel with the item column, data_value, and (when scale is not NULL) scale_id.

key

Name of the key column for the default path. Ignored on the items/agg path, where the measure is keyed on item.

score

Name of the numeric score column for the default path. Ignored on the items/agg path, where the score is the data_value on scale.

key_type

Measure grain: occupation, task, or DWA. Forced to "task" on the items/agg path.

universe

Optional vector or data frame of valid keys.

measure_id

Short identifier for the measure.

measure_name

Human-readable measure name.

source

Optional source label.

release_version

Optional O*NET release used to create the measure.

weight_panel

Optional weight panel used to report employment coverage for occupation-level measures.

items

Optional character vector of target task ids. Supplying it (or agg) selects the convenience path. Required when agg = "targeted".

agg

Aggregation mode for the convenience path. "targeted" restricts the panel to items; "aggregate" keeps every item. Defaults to "targeted" when items is supplied.

item

Column identifying the content item on the items/agg path and used as the measure key. Defaults to "task_id".

scale

Scale id used to select one rating row per task on the items/agg path, for example "IM" for Importance. Use NULL to keep every row of data, which requires item to be unique on its own.

Value

An onet_measure object with data, coverage, unmatched, and metadata fields. The items/agg path returns a task-grain measure (key_type = "task") keyed on item and scored on the scale rating, ready for onet_task_to_occupation().

Details

Two construction paths are available. The default path validates a table of key and score columns you already built. The convenience path, selected by passing items or agg, builds a task-grain measure in one line from a Task Ratings style panel: it restricts the panel to the caller's target items (when agg = "targeted") or keeps every item (when agg = "aggregate"), selects one rating row per task on scale (default Importance, "IM"), and keys the result on item (default "task_id"). The result is exactly the measure the default path returns on that same subset, so it is a thin convenience wrapper, not a new estimator. Roll it up to occupations with onet_task_to_occupation() and onet_measure_aggregate(). The target items, not the package, carry the substantive judgement.

A single items vector expresses one target set, so the switch builds a targeted composite such as a routine-task composite. A difference index like routine-task intensity, which subtracts abstract and manual composites from a routine composite, is composed separately from two or more such measures.

Examples

scores <- tibble::tibble(
  onet_soc_code = c("15-1252.00", "29-1141.00"),
  score = c(0.7, 0.2)
)
universe <- c("15-1252.00", "29-1141.00", "11-1011.00")
measure <- onet_measure(scores, "onet_soc_code", "score", universe = universe)
onet_measure_coverage(measure)
#> # A tibble: 1 × 6
#>   key_type   n_input n_universe n_matched coverage_share employment_coverage_s…¹
#>   <chr>        <int>      <int>     <int>          <dbl>                   <dbl>
#> 1 occupation       2          3         2          0.667                      NA
#> # ℹ abbreviated name: ¹​employment_coverage_share

# Convenience path: build a task-grain targeted measure from a panel in one
# line, ready to roll up with onet_task_to_occupation().
panel <- tibble::tibble(
  onet_soc_code = rep(c("15-1252.00", "29-1141.00"), each = 3),
  task_id = c("1", "2", "3", "4", "5", "6"),
  scale_id = "IM",
  data_value = c(4.5, 3.0, 2.0, 1.0, 4.0, 3.0)
)
targeted <- onet_measure(panel, items = c("1", "5"), agg = "targeted")
targeted$data
#> # A tibble: 2 × 6
#>   onet_soc_code task_id scale_id data_value measure_key measure_score
#>   <chr>         <chr>   <chr>         <dbl> <chr>               <dbl>
#> 1 15-1252.00    1       IM              4.5 1                     4.5
#> 2 29-1141.00    5       IM              4   5                     4