Single Answer Questions Research Articles

Comparing The Effectiveness Of Multiple-Answer And Single-Answer Multiple-Choice Questions In Assessing Student Learning

ABSTRACT Driven by technological innovations, continuous digital expansion has transformed fundamentally the landscape of modern higher education, leading to discussions about evaluation techniques. The emergence of generative artificial intelligence raises questions about reliability and academic honesty regarding multiple-choice assessments in online education. In this context, this study investigates multiple-answer questions (MAQs) versus traditional single-answer questions (SAQs) in online higher-education assessments. A mixed-methods study involving quantitative field experiments and qualitative interviews was conducted with students enrolled in an online Marketing M.Sc. program. The students were divided randomly and assessed using either SAQs or MAQs, and the impacts on test performance of variables such as grade averages, study times, perceived workload, and difficulty were evaluated using independent-sample t-tests and ordinary least-squares regression analysis. The results show that although grades were lower and MAQs were perceived as being more difficult, study times and perceived workload did not differ significantly between the two formats. These findings suggest that despite their challenge, MAQs can promote deeper understanding and greater learning retention. Furthermore, even with their higher perceived difficulty and impact on performance, MAQs hold potential for dealing with academic-integrity concerns related to artificial intelligence.

Performance evaluation of ChatGPT, GPT-4, and Bard on the official board examination of the Japan Radiology Society.

Herein, we assessed the accuracy of large language models (LLMs) in generating responses to questions in clinical radiology practice. We compared the performance of ChatGPT, GPT-4, and Google Bard using questions from the Japan Radiology Board Examination (JRBE). In total, 103 questions from the JRBE 2022 were used with permission from the Japan Radiological Society. These questions were categorized by pattern, required level of thinking, and topic. McNemar's test was used to compare the proportion of correct responses between the LLMs. Fisher's exact test was used to assess the performance of GPT-4 for each topic category. ChatGPT, GPT-4, and Google Bard correctly answered 40.8% (42 of 103), 65.0% (67 of 103), and 38.8% (40 of 103) of the questions, respectively. GPT-4 significantly outperformed ChatGPT by 24.2% (p < 0.001) and Google Bard by 26.2% (p < 0.001). In the categorical analysis by level of thinking, GPT-4 correctly answered 79.7% of the lower-order questions, which was significantly higher than ChatGPT or Google Bard (p < 0.001). The categorical analysis by question pattern revealed GPT-4's superiority over ChatGPT (67.4% vs. 46.5%, p = 0.004) and Google Bard (39.5%, p < 0.001) in the single-answer questions. The categorical analysis by topic revealed that GPT-4 outperformed ChatGPT (40%, p = 0.013) and Google Bard (26.7%, p = 0.004). No significant differences were observed between the LLMs in the categories not mentioned above. The performance of GPT-4 was significantly better in nuclear medicine (93.3%) than in diagnostic radiology (55.8%; p < 0.001). GPT-4 also performed better on lower-order questions than on higher-order questions (79.7% vs. 45.5%, p < 0.001). ChatGPTplus based on GPT-4 scored 65% when answering Japanese questions from the JRBE, outperforming ChatGPT and Google Bard. This highlights the potential of using LLMs to address advanced clinical questions in the field of radiology in Japan.

Assessment of the effectiveness of BOPPPS-based hybrid teaching model in physiology education

BackgroundOnline teaching has become increasingly common in higher education of the post-pandemic era. While a traditional face-to-face lecture or offline teaching remains very important and necessary for students to learn the medical knowledge systematically, guided by the BOPPPS teaching model, combination of online and offline learning approaches has become an unavoidable trend for maximizing teaching efficiency. However, in physiological education, the effectiveness of combined online teaching and offline teaching models remains poorly assessed. The present study aims at providing an assessment to the hybrid teaching model.MethodsThe study was performed among undergraduate medical students of Class 2017 ~ 2019 in the Physiology course in Harbin Medical University during 2018-2020. Based on established offline teaching model with BOPPPS components in 2018, we incorporated online teaching contents into it to form a hybrid BOPPPS teaching model (HBOPPPS, in brief), preliminarily in 2019 and completely in 2020. HBOPPPS effectiveness was assessed through comparing the final examination scores of both objective (multi-choice and single answer questions) and subjective (short and long essays) questions between classes taught with different modalities.ResultsThe final examination score of students in Class 2019 (83.9 ± 0.5) who were taught with the HBOPPPS was significantly higher than that in Class 2017 (81.1 ± 0.6) taught with offline BOPPPS and in Class 2018 (82.0 ± 0.5) taught with immature HBOPPPS. The difference mainly attributed to the increase in average subjective scores (41.6 ± 0.3 in Class 2019, 41.4 ± 0.3 in Class 2018, and 38.2 ± 0.4 in Class 2017). In the questionnaire about the HBOPPPS among students in Class 2019, 86.2% responded positively and 79.4% perceived improvement in their learning ability. In addition, 73.5% of the students appreciated the reproducibility of learning content and 54.2% valued the flexibility of HBOPPPS. Lastly, 61.7% of the students preferred the HBOPPPS relative to BOPPPS in future learning.ConclusionsHBOPPPS is likely a more effective teaching model and useful for enhancing effectiveness of Physiology teaching. This is attributable to the reproducibility and flexibility as well as the increased learning initiatives.

사상체질에 따른 식습관과 주관적 건강상태와의 연관성

Dietary habits are known to be closely related not only to chronic diseases such as cardiovascular disease and diabetes, but also to self-rated health (SRH). Consequently, there is a highly important lifestyle component in maintaining a healthy life. The aim of the present study was to determine the correlation between diverse dietary habits based on the Sasang constitutions (SC) and SRH. The data were collected from 2,984 participants who met the constitution pharmacology criteria and whose SC were clinically confirmed by SC professionals. Dietary habits were observed by assessing the regularity of diet, meal volume, and meal speed; single-answer questions were used for SRH. Fisher’s exact/chi-square test and logistic regression were used to analyze the correlation between dietary habits depending on constitutional type and SRH. In association with constitutional SRH, the Taeum type is related to the amount of meal, the Soeum and Soyang types are related to the regularity of meal and meal volume. In the Taeum type, there was a significantly higher risk of lower SRH when meal volumes were irregular less than previous amounts. Irregular diet, as well as increased or decreased meal volume, negatively correlated with SRH in the Soeum type; in the Soyang type, irregular diet and decreased or irregular meal volume negatively affected SRH. This showed the presence of the different effects of dietary habits on health depending on constitution and also confirmed the association of the regularity of diet and meal volume based on constitution and the frequency of bad dietary habits with SRH. We suggest that personalized healthcare with adequate dietary improvement considering constitution is needed for health management or improvement.

G315(P) Doctors, death and diversity: are we aware of the cultural needs of parents during bereavement care?

Background When a child dies, the interactions between medical professionals and parents are remembered clearly and impact significantly both positively and negatively on the family’s grieving process and eventual recovery. In our richly multi-cultural society it is an increasing challenge to provide sensitive, appropriate bereavement care. Aims To determine whether paediatric doctors have basic knowledge of the cultural practices during end of life care associated with the five main religious affiliations present in our community (Muslim, Christian, Jewish, Hindu and No Religion). Methods An electronic questionnaire was sent to all paediatric trainees and consultants working in hospitals in one training sector of the UK. A combination of case based multiple choice questions and single answer questions were used. Results were collated anonymously. Results 111 (28%) of questionnaires were returned. Most doctors were able to identify Faith Leaders of Islam and Judaism (83% and 81% respectively), and identify support for those of no specific religious affiliation, however only 5% of respondents could name a Faith Leader for Hinduism. Regarding actions at the time of death; Christian rituals were best understood (66%) followed by Islamic (54%), Hindu (44%) and Jewish (30%) traditions. Only 28% of respondents were aware that organ donation is not strictly forbidden in any of these religions. Overall, there was no relationship between training grade and cultural knowledge. Conclusions As doctors we are not well informed about the predominant religious needs of our community. This is likely to impact adversely on our ability to offer appropriate bereavement care to parents. Our false beliefs about barriers to organ donation may also prevent us from offering this opportunity to newly bereaved parents. We plan to produce a nationally available training package in order to equip paediatric doctors with the knowledge to provide high quality, culturally sensitive end of life care.

Nonrestricted multiple‐choice examination items

Multiple-choice items are frequently used in objective examinations. The format chosen should conform to the nature of the instruction. Knowledge about cumulative information, such as lists of attributes, can be tested efficiently by means of multiple-choice items that include a variable number of correct answers. In contrast to conventional, single-answer questions, nonrestricted multiple-choice items are capable of including more facts and fewer incorrect responses. In addition, the nonrestricted format is not burdened with the repetitious pattern of one correct answer coupled with several incorrect responses, a cue that may promote successful guessing. Item analyses can be performed on examinations that include both conventional and nonrestricted items. The reliability of one examination constructed totally with nonrestricted items was analyzed by means of the Kuder-Richardson Formula No. 20. The value 0.72 proved this examination to be both discriminating and consistent.