Guessing in multiple-choice assessments threatens the validity of achievement scores, particularly in secondary school Economics where such formats are widely used. This study investigated strategies to minimize pseudo-guessing in SS2 Economics multiple-choice tests using the Three-Parameter Logistic (3PL) model within the Item Response Theory framework. The study adopted an instrumental/measurement research design. The population comprised 2,266 SS2 Economics students in government-owned secondary schools in Nsukka Education Zone, Enugu State, Nigeria. A sample of 340 students was selected using Yamane’s formula and stratified proportionate random sampling. A 40-item multiple-choice Economics test was developed, face and content validated by experts, and pilot tested prior to administration. The instrument was administered as a pre-test and post-test, and data were analyzed using the R statistical environment (mirt package), focusing on the pseudo-guessing parameter (c). Pre-test results revealed c values ranging from 0.18 to 0.25, indicating moderate susceptibility to guessing. Following the implementation of strategies such as improved distractor plausibility and clearer instructions, post-test c values decreased to a range of 0.08 to 0.14. The reduction values (0.09–0.11 across items) demonstrated a consistent decline in guessing probability. The study concludes that targeted item construction and psychometric calibration significantly reduce pseudo-guessing and enhance the validity of multiple-choice assessments in secondary school Economics.
Ossai Elizabeth Ngozika (Wed,) studied this question.