Dexter
Dexter is an R package for psychometric analysis of data from
educational and psychological tests. Dexter typically works with project
database files saved on disk.
Installation
install.packages('dexter')
If you encounter a bug, please post a minimal reproducible example on
github. We post
news and examples on a website and
blog.
Example
library(dexter)
# start a project and fill it with data
# verbAggrRules and verbAggrData are example datasets provided with dexter
db = start_new_project(verbAggrRules, "verbAggression.db")
add_booklet(db, verbAggrData, booklet_id = "verb_agg")
# Classical test theory
tia = tia_tables(db)
tia$booklets
head(tia$items)
| booklet_id | item_id | mean_score | sd_score | max_score | pvalue | rit | rir | n_persons |
|---|
| verb_agg | S1DoCurse | 1.082 | 0.807 | 2 | 0.541 | 0.582 | 0.519 | 316 |
| verb_agg | S1DoScold | 0.832 | 0.815 | 2 | 0.416 | 0.651 | 0.596 | 316 |
| verb_agg | S1DoShout | 0.468 | 0.709 | 2 | 0.234 | 0.520 | 0.460 | 316 |
| verb_agg | S1WantCurse | 1.123 | 0.827 | 2 | 0.562 | 0.537 | 0.468 | 316 |
| verb_agg | S1WantScold | 0.930 | 0.850 | 2 | 0.465 | 0.593 | 0.528 | 316 |
| verb_agg | S1WantShout | 0.712 | 0.777 | 2 | 0.356 | 0.529 | 0.464 | 316 |
# IRT, extended nominal response model
f = fit_enorm(db)
head(coef(f))
| item_id | item_score | beta | SE_beta |
|---|
| S1DoCurse | 1 | -1.3422140 | 0.1541565 |
| S1DoCurse | 2 | -0.6375015 | 0.1418423 |
| S1DoScold | 1 | -0.6702036 | 0.1429057 |
| S1DoScold | 2 | -0.2589855 | 0.1579467 |
| S1DoShout | 1 | 0.3254326 | 0.1480166 |
| S1DoShout | 2 | 0.3687574 | 0.2099654 |
# ability estimates per person
abl = ability(db, parms = f)
head(abl)
| booklet_id | person_id | booklet_score | theta |
|---|
| verb_agg | dx_0000001 | 13 | -1.0238738 |
| verb_agg | dx_0000002 | 28 | 0.3124831 |
| verb_agg | dx_0000003 | 4 | -2.3748882 |
| verb_agg | dx_0000004 | 19 | -0.4630604 |
| verb_agg | dx_0000005 | 7 | -1.7721275 |
| verb_agg | dx_0000006 | 25 | 0.0512826 |
# ability estimates without item S1DoScold
abl2 = ability(db, parms = f, item_id != "S1DoScold")
# plausible values
pv = plausible_values(db, parms = f, nPV = 5)
head(pv)
| booklet_id | person_id | booklet_score | PV1 | PV2 | PV3 | PV4 | PV5 |
|---|
| verb_agg | dx_0000001 | 13 | -1.1129358 | -0.3718473 | -0.6251660 | -1.0861430 | -1.4627665 |
| verb_agg | dx_0000002 | 28 | -0.0643628 | 0.1312981 | -0.6255087 | -0.1063878 | 0.3527158 |
| verb_agg | dx_0000003 | 4 | -1.8603752 | -2.5817701 | -1.6376942 | -1.7172643 | -1.8129040 |
| verb_agg | dx_0000004 | 19 | -1.0663677 | -0.6810231 | -1.1512542 | -0.9481833 | -0.8004328 |
| verb_agg | dx_0000005 | 7 | -2.0083386 | -1.5279208 | -1.3307093 | -1.5430697 | -1.3224421 |
| verb_agg | dx_0000006 | 25 | 0.2020330 | 0.2050144 | 0.2221347 | -0.4249511 | 0.0809038 |