Locates every item (and every score category of a polytomous item) on the
ability scale at the response probability rp: the ability at which a test
taker has probability rp of answering correctly (or of reaching that
category or higher). Sorted by location, this is the ordered item booklet
of the Bookmark method; a panelist's bookmark translates to a cut at the
location of the bookmarked item.
Arguments
- p
a calibrated pipeline.
- rp
response probability, commonly 0.67 (or 0.50).
- form
optional fixed form number from
assemble_forms().
Value
A tibble with one row per item category: page (position in the
booklet), item, category, and theta, the location at rp. Items
that never reach rp (a lower asymptote above it) have NA and come
last.
Examples
# \donttest{
p <- calibrate(add_responses(mv_pipeline(), Science), model = "graded")
#> Warning: Calibration is running with 1 review open flag(s) from earlier stages:
#> #1 [review] sparse_category: Comfort (category 1: n = 5)
#> Flags only warn because strict = FALSE. See flags(p).
item_map(p)
#> # A tibble: 12 × 4
#> page item category theta
#> <int> <chr> <int> <dbl>
#> 1 1 Comfort 1 -3.99
#> 2 2 Benefit 1 -2.41
#> 3 3 Future 1 -1.97
#> 4 4 Work 1 -1.81
#> 5 5 Comfort 2 NA
#> 6 6 Comfort 3 NA
#> 7 7 Work 2 NA
#> 8 8 Work 3 NA
#> 9 9 Future 2 NA
#> 10 10 Future 3 NA
#> 11 11 Benefit 2 NA
#> 12 12 Benefit 3 NA
# }