Skip to contents

The ready-made sections that new_lab() and add_lab_section() use. Each works with any dataset: Its numbers, checks and questions are computed from the lab's data.

Usage

lab_section_types()

Value

A tibble with the section type, its title, a description, and needs: What the section needs besides the responses ("group", "item text", "item families" or "scores"), or "".

Details

The last three sections teach AI in measurement with the lab's data and need more than responses (see the item_info and scores arguments of new_lab()): "item-features" predicts item difficulty from the item text, "aig" checks whether the items of each family are equally difficult, and "ai-scoring" audits machine scores against human scores.

Examples

lab_section_types()
#> # A tibble: 13 × 4
#>    type           title                           description              needs
#>    <chr>          <chr>                           <chr>                    <chr>
#>  1 intro          Meet the data                   The story, the size of … ""   
#>  2 classical      Classical item statistics       Item means and item-tot… ""   
#>  3 dimensionality How many dimensions?            Compare one- and two-fa… ""   
#>  4 models         Compare IRT models              Compare the Rasch model… ""   
#>  5 item-fit       Item fit                        S-X2 item fit with item… ""   
#>  6 information    Information and precision       Test information and st… ""   
#>  7 scoring        Scoring people                  EAP scores with fscores… ""   
#>  8 dif            Are the items fair?             Likelihood-ratio DIF te… "gro…
#>  9 reflection     Your recommendation             A written recommendatio… ""   
#> 10 wrap-up        Where to next                   A summary and pointers … ""   
#> 11 item-features  What the item text tells us     Text features with item… "ite…
#> 12 aig            Item families                   Are the items of each f… "ite…
#> 13 ai-scoring     Machine scores and human scores Machine-human agreement… "sco…