Abstract
The study ascertained the impact of number of options on distracter performance when confidence scoring was used and established the impact of numbers of options for multiple choice test items on the reliability of the test. The survey research design was adopted for the study and the population consisted secondary school students in Osun State. The sample comprised 360 students randomly selected from 18 schools. The instrument used for the study was 2015 West Africa Scoool Certificate Examination (WAEC) Economics test items. The Economics Achievement test (EAT) which was the 4-options format adapted from the 2015 WASCE Economics items had been validated for used by the West African Examination Council. The validity of the other types (3-options and 5-options format) were determine and scrutinized by experts in Test and Measurement and Economics teachers in the secondary school to judge its face and content as well as item arrangement. The corrections were incorporated into final version of the instruments. The EAT instrument of 3-options and 5-options were validated in a pilot study conducted using 40 senior secondary school II Economics students who were not part of the final sample size but in a different study area with similar characteristics. Given the responses of the respondents used in the pilot study, the 3-options and 5-options which consisted of 50 items were subjected to a measure of internal consistency using Kuder-Richardson 21 (KR21) to ascertain the reliability of the instrument. The result of the K-21 for both 3-options and 5-options yielded coefficient of 0.79 and 0.83 respectively. Data collected were analysed using ANOVA, Kuder-Richarson Formula (KR-20) and Fisher’s Z-Test with aid of FZT compotator. The results of the study showed that number of options had significant impact on distracters’ performance when scored using confidence scoring (F= 6.679,p <0.05). The results also showed that for each pair wise comparison of 3/4-options (Z =0.640), 3/5-options (Z=0.196) at p<0.05 the difference in the reliability coefficient were not significant. The study therefore concluded that option length of multiple choice objective test item have impact on its reliability and distracter efficiency.