Discriminant Validity and Reliability of Scores of Multiple Choice and Short Essay Questions

Abstract

Background and Objective: Assessment of learning plays a vital role in curriculum allowing faculty to evaluate the achievement of student and educational process. Quality assurance of the assessments is done to assess the quality of teaching and learning. A study is therefore conducted to find the discriminant validity and reliability of scores of Multiple Choices and Short Answer Questions.Methods: A correlation study was conducted on 272 second year MBBS students of Fatima Memorial College of Medicine and Dentistry, Lahore from 2016 -2017. Students were assessed in end of year examination with multiple choice questions (MCQs) and short essay questions and (SEQs) in the subjects of Physiology and Biochemistry. A total of 149 second-year MBBS students took thePhysiologyand143 studentstook the end of year Biochemistry examination with 50% minimum passing level. The mean and standard deviation of the scores were calculated and the scores of MCQs and SEQs were correlated by applying Pearson's correlation. Reliability was determined by Cronbach's alpha. Discriminant validity of scores (MCQS and SEQS) was analyzed by Pearson's correlation.Results: Students scored significantly higher in MCQsthan SEQsin end of year examination of Biochemistry and Physiology. Positive significant correlation of scores (MCQs and SEQs) was obtained in Biochemistry while correlation was positive but non-significant in Physiology. The reliability assessed by Cronbach's alpha was moderate in Biochemistry and Physiology assessments. Discriminant validity was determined in both subjects by determining the correlation between the scores of the MCQs and SEQs, constructed according to the same module. Conclusion:The internal assessment of students based on end of year examinations in the subjects of Physiology and Biochemistry scored moderate in terms of reliability and discriminant validity was found between scores of MCQs and SEQs in each module.

Discrimination Power of Short Essay Questions Versus Multiple Choice Questions as an Assessment Tool in Clinical Biochemistry

“…It is more discriminating than multiple questions. Our study's results are in agreement with those in the studies by Jaleel et al (2020) [17], Kunjappagounder et al [16], and Maryani et al [18].…”

“…As expected, the present study revealed positive linear correlations between SEQs, MCQs, and total grades. These correlations were also reported in other subjects, such as pharmacology [25], otorhinolaryngology [26], pediatrics [27], and basic medical sciences, including physiology and medical biochemistry [4,17].…”

“…Moreover, SEQs necessitate students to interpret and analyze their thoughts to provide the answers; they also evaluate written communication skills [ 17 ].…”

Item analysis: the impact of distractor efficiency on the difficulty index and discrimination power of multiple-choice items

“…High-quality MCQs are considered an appropriate assessment tool because they can cover a wide range of knowledge and domains of knowledge [ 1 ]. Many authors reported the validity and reliability of MCQs [ 2 – 6 ]. The validity and reliability of MCQs can be ensured pre-construction by the presence of the content material from which the MCQs will be constructed and a blueprint [ 7 – 10 ].…”