A Psychometric Model for Discrete-Option Multiple-Choice Items.

Area of Science:

• Educational Measurement
• Psychometrics
• Cognitive Psychology

Background:

• Traditional multiple-choice (MC) items present all options simultaneously.
• Discrete-option multiple-choice (DOMC) items reveal options sequentially, enhancing item security and reducing testwiseness.
• DOMC's random key location can influence item difficulty.

Purpose of the Study:

• To propose a psychometric model for DOMC items that accounts for random key location.
• To investigate the impact of key location variability on DOMC item difficulty and person proficiencies.
• To differentiate sources of item difficulty and person abilities within the DOMC format.

Main Methods:

• Development of a psychometric model for DOMC items.
• Analysis of two empirical datasets with items in both DOMC and MC formats.
• Examination of key location effects across items and persons.

Main Results:

• The proposed model effectively isolates distinct sources of item difficulty (keyed vs. distractor options).
• The model differentiates distinct person proficiencies related to these difficulty components.
• Variability in key location significantly affects DOMC item difficulty.

Conclusions:

• DOMC items offer a unique opportunity to dissect item difficulty and person abilities.
• The psychometric model provides valuable insights into DOMC item functioning.
• Findings have practical implications for designing DOMC test administrations with randomized key locations.