Numerical-Reasoning Test (Short Form)
Reliability & validity
Reliability & validity evidence is syncing from Principia.
Sources
- Baron and Austin
Deploy it
The full item bank and survey deployment live in the People Analytics Toolbox (the enterprise survey-item library), with the min-N privacy gate.
What the literature says
The measurement literature behind this signal — sourced, so you can defend it.
“Perhaps more significantly, the price of these screens is declining to within a range that makes them practical for use in large-scale testing programs. Multiple items for same passage can lead to violations of the IRT assumption of local independence, which can yield…”
— computerized_adaptive_testingmatch 55%
“The results of the factor analyses for PC, QR, and SK are summarized in Tables 8.6 , 8.7 and 8.8 , respectively. The tables present (a) the proportion of correct responses, (b) the guessing level parameter, (c) a location or difficulty parameter, (d) the (PROMAX-rotated) factor…”
— computerized_adaptive_testingmatch 53%
“42 of random number, 120 Reference estimates, 94 Regressed test scores, 164 Regression, 150 Regression coefficients, 201 , 211 –212 Relationship between equating samples and population, 145 –146 between item content and item selection procedures, 187 between old and new forms,…”
— computerized_adaptive_testingmatch 50%
Resources: computerized_adaptive_testing