Computerization and adaptive administration of the NEO PI-R.

Area of Science:

• Psychometrics
• Personality Assessment
• Psychological Measurement

Background:

• The NEO Personality Inventory-Revised (NEO PI-R) is a widely used measure of personality.
• Computerized adaptive testing (CAT) offers potential for efficient assessment.
• Item response theory (IRT) is a statistical framework for developing adaptive tests.

Purpose of the Study:

• To evaluate the effectiveness of an IRT-based CAT algorithm for the NEO PI-R.
• To determine the precision of personality facet scores obtained via CAT.
• To compare the efficiency of CAT with a fixed-item approach.

Main Methods:

• Real-data simulations were conducted with a sample size of 1,059 participants.
• A maximum information item selection algorithm was employed for the CAT.
• The recovery of full-scale facet scores was assessed.

Main Results:

• Satisfactory recovery of NEO PI-R facet scores was achieved with approximately four items per scale.
• CAT administration could potentially halve the length of the NEO PI-R with minimal loss in precision.
• The necessity of a complex CAT algorithm was questioned, as a fixed set of 'best' items often yielded comparable results.

Conclusions:

• IRT-based CAT is a viable method for efficient personality assessment with the NEO PI-R.
• A simpler, non-adaptive selection of high-quality items may offer similar psychometric benefits.
• The future of computerized personality assessment may involve IRT methods for optimized measurement.