PSYCHOMETRIC VALIDATION OF THE POLICE SELECTION INSTRUMENT USING AN ORTHOGONAL MULTIDIMENSIONAL RASCH MODEL (MRM)
DOI:
https://doi.org/10.7492/qzzm5s79Keywords:
multidimensional measurement, specialized competency assessment, personnel selection, item response theory, differential item functioningAbstract
Traditional assessment tools for specialized police units often collapse distinct competencies into a single composite score, obscuring the
operationally critical differences among specialist roles. This study challenges that assumption by introducing a psychometric framework that
treats counter-terrorism, bomb disposal, chemical-biological-radiological-nuclear (CBRN) threat management, and technical support as
fundamentally independent measurement dimensions. A sixty-item selection instrument was developed for Detachment Gegana of the Indonesian
National Police and validated using the Orthogonal Multidimensional Rasch Model (OMRM) with data from 321 active officer candidates across
twelve regional units. A sequential validation framework integrated exploratory factor analysis, confirmatory factor analysis, model selection,
item fit assessment, reliability estimation, and differential item functioning analysis. The orthogonal four-factor confirmatory model
demonstrated excellent fit (???????????????????? = 0.005, ???????????? = 0.990, ???????????? = 0.990), with a non-significant chi-square difference test (????????² =
0.834, ???????? = 6, ???? = 0.991) confirming that expertise in one domain carries virtually no predictive value for another. The Multidimensional
Rasch Model was preferred over its two-parameter alternative based on substantially superior parsimony-adjusted fit (???????????????? = 271.70). All
sixty items met acceptable fit criteria (RMSEA: 0.000– 0.095), item reliability was uniformly excellent across domains (range: 0.979– 0.987),
and no item exhibited significant differential item functioning after false discovery rate correction. These findings establish that effective selection
for specialized roles requires measuring each competency independently, offering a validated and practically actionable framework for evidencebased personnel assignment in elite law enforcement units.
