Difficulty Of Questions Research Articles

Quality of Tests on Mathematics Questions for Year-End Assessment (PAT) Class VIII at SMPN 2 Bangkinang Kota

This study aims to determine the quality of the test on multiple choice questions in mathematics for Year End Assessment (PAT) class VIII SMPN 2 Bangkinang Kota for the 2019/2020 academic year based on the level of question validity, reliability level, difficulty level, distinguishing power, and option effectiveness cheater). This research is an evaluation with a quantitative descriptive approach. The research was conducted with a sample of 30 student answer sheets. Techniques for collecting data are documentation techniques. After the data is collected, the data will be analyzed with the help of Excel. Judging from the results of this analysis, it can be seen that the End of Year Assessment (PAT) questions for class VIII SMPN 2 Bangkinang Kota for the 2019/2020 academic year were not carried out by quantitative reviewers. Judging from the level of validity, it consists of 14 valid questions, while the other 16 questions are said to be invalid. The PAT questions have a high level of reliability. As for the difficulty level of the questions, there are 7 items (23.33%) difficult, 18 items (60%) moderate, and 5 items (16.66%) easy. The distinguishing power is 8 questions that are bad (26.66%), 7 questions with sufficient criteria (23.33%), 7 questions with good criteria (23.33%), and 8 questions with very good criteria (26.66%) ). For the effectiveness of the option, 4 items (13.33%) were very good, 15 items (50%) were good, 9 items (30%) were bad, and 2 items (6.33%) were very bad. So that based on the results of the quantitative test, it turns out that most of the items were not of good quality for Class VIII Year End Assessment (PAT) for the 2019/2020 school year.

Catquest-9SF functioning over a decade \u2013 a study from the Swedish National Cataract Register

BackgroundThe Swedish National Cataract Register (NCR) collects data on cataract surgery outcomes during March, including patient-reported outcomes using the Catquest-9SF questionnaire for over 11 years. Previous studies from NCR have shown that the preoperative visual acuity has improved over time. The main purpose of this study was to evaluate the Catquest-9SF Rasch scoring performance in this changing environment. A second purpose was to describe clinical data over the same period for those who completed the questionnaire.MethodsThe performance of the Catquest-9SF was analysed by a separate Rasch analysis for each year, resulting in a preoperative and postoperative score for each participating patient in the annual cohorts. The clinical data and questionnaire scoring were analysed for each year in the period 2008–2018 inclusive.ResultsData were available for 42,023 eyes for 11 annual cohorts (2008–2018). The psychometric properties of the questionnaire were stable during the study period. Person separation (precision) for the whole period was 2.58 and varied between 2.45 and 2.72. The person reliability was 0.87 and varied between 0.86 and 0.88. The targeting of question difficulty to person ability became less accurate over time meaning that the item activities became easier to carry out without difficulty. The average targeting for the whole period was −2.06 and changed from −1.92 in 2008 to −2.31 in 2018. The person score improved both before surgery and after surgery, indicating that patients are undergoing surgery at a more able level and getting better outcomes. The average improvement by surgery decreased from 3.41 logits in 2008 to 3.21 logits in 2018 (p = 0.003).Over time, patient age decreased from 75 to 74 years (p < 0.001) and the proportion of women decreased from 63.9 to 57.9% (p < 0.001). The mean preoperative visual acuity in both the operated eye and the better eye improved over time (0.47 to 0.40 logMAR, p < 0.001 and 0.22 to 0.19 logMAR, p < 0.001, respectively), as did the mean postoperative visual acuity in the operated eye (0.14 to 0.09 logMAR, p < 0.001).ConclusionsThe Catquest-9SF retained stable psychometric properties over this 11-year period although more recent cohorts included slightly younger patients with somewhat better vision.

Logits-based processing of educational information

The objective of this work is to study the processing of educational information using logits. The one-parameter Rasch model is used as a basis for test data simple processing. We consider the statistics on a group of students and on different groups. Based on the test results, we consider two situations and two correlations. The first correlation is between grades and knowledge. The second one is grade dependence on the question difficulty. Two parameters are evaluated: the knowledge level and the question difficulty. We describe each situation in a diagram form. The model parameters are discussed. Formulas for calculation are given. The paper shows a mechanism for analyzing the results and how qualimetric evaluations are connected with the real knowledge level and test complexity.

Analysis the Difficulty of Indonesian Language Class V Elementary School / Analisis Tingkat Kesulitan Bahasa Indonesia Kelas V Madrasah Ibtidaiyah

Indonesian Language subject is often considered easy by studensts, students are able to answer questions even though they don’t study it, and get the highest score compared to other subjects. This study aims to explain the level of difficulty and cognitive level of question of Bahasa Indonesia subject in Class V MIN 2 Banda Aceh. The question were given in the examination of odd semester 2019. The type of this research is quantitative descriptive by examining the relationship between one variable with another variable. Data collection techniques to get more detailed information wasgained by interviewing Teacher of Bahasa Indonesia Subject of Class V MIN 2 Banda Aceh. Object of this study was question sheet of Bahasa Indonesia in semester 2019, and student answers. The results showed that the difficulty level of the questions was 25% in the easy category, 65% the medium category, and 10% in the difficult categories. On the other hand, the cognitive level of questions is87.5% in the categories C1, 5% in the categories C2 questions and 7.75% in the categories C6 questions while C3, C4, and C5 categories are not found.

Students' difficulties in similar triangle questions

Similar triangles in questions are usually given as separate, adjacent or overlapped. Furthermore, similarity types such as Side-Angle-Side (S.A.S.), Side-Side-Side (S.S.S.) and Angle-Angle (A.A.) are requested in the questions. Students have more trouble in other types of questions. The purpose of this study is to investigate the difficulties of students about similar triangles and the reasons for these difficulties. This research was carried out with the case study method, which is one of the qualitative research approaches. The study was conducted with 55 Science High School 9th grade students and 9 open-ended questions were used to examine students' knowledge about “similarity in triangles”. Furthermore, 5 students were interviewed to find out the reasons for their solutions. Descriptive analysis method was used to analyze the data. As a result, it can be concluded that students have difficulties mostly in overlapped triangles and Angle-Angle type questions. On the other hand, it can be concluded that students are quite successful where similar triangles are given separately. In the light of the findings obtained in this study, it can be advised for lecturers to focus on the questions where similar triangles are overlapped while explaining the similarity in the triangle.&#x0D; Keywords: Similarity, Triangles, Difficulties, High School Students.

Automatic Grading and Adaptive Question Selection for English Language Testing

Computer based evaluation of students’ performance is playing a vital role worldwide for ELP examinations. This method is more efficient than manual evaluating process. In this study, a new method is proposed to evaluate the examinees descriptive answers using Text Mining and Text processing algorithms with the help of Natural Language Processing tools. This system includes three sections of ELP, namely: reading, writing and speaking. Dataset for the examination process is extracted using the UiPath platform from reference books which are stored into the database. On taking a test, the system automatically calculates the result and polarity of the subjective answers using NLP and fuzzy logic. In this study, we used Rasch algorithm for adapting to the ability of the examinee for question difficulty level selection and also evaluate the same answer using Natural language processing (NLP) algorithm to check grammar mistakes, polarity and stores the marks in database and finally represent them graphically for better visualization of results.

Case Study: Students’ Code-Tracing Skills and Calibration of Questions for Computer Adaptive Tests

Computer adaptive testing (CAT) enables an individualization of tests and better accuracy of knowledge level determination. In CAT, all test participants receive a uniquely tailored set of questions. The number and the difficulty of the next question depend on whether the respondent’s previous answer was correct or incorrect. In order for CAT to work properly, it needs questions with suitably defined levels of difficulty. In this work, the authors compare the results of questions’ difficulty determination given by experts (teachers) and students. Bachelor students of informatics in their first, second, and third year of studies at Subotica Tech—College of Applied Sciences had to answer 44 programming questions in a test and estimate the difficulty for each of those questions. Analyzing the correct answers shows that the basic programming knowledge, taught in the first year of study, evolves very slowly among senior students. The comparison of estimations on questions difficulty highlights that the senior students have a better understanding of basic programming tasks; thus, their estimation of difficulty approximates to that given by the experts.

Preservice mathematics teachers’ competencies in the process of transformation between representations for the concept of limit: A qualitative study

In this study, external representations and the problems encountered related transformation process between representations towards limit concept were investigated. "Limit Representation Conversion Test" was administered to 41 preservice mathematics teachers studying at a state university in central Turkey during 2018–2019 academic years. In this study, which was designed with the case study model, which is one of the qualitative research models, the data were analyzed by content analysis. Unstructured interviews were made with preservice mathematics teachers whose explanations were insufficient or differed and the problems encountered were determined. It was observed that preservice mathematics teachers had most difficulties in the verbal representation type questions. It was revealed that preservice mathematics teachers who gave the wrong answers mostly had deficiencies in the concept and the process and could not fully understand the limit problems. It was determined that preservice mathematics teachers had difficulties in knowing the concept of limit point, determining the function and interpreting verbal data. It was seen that preservice mathematics teachers who proceeded towards the concept and process answered wrong due to mathematical operations errors and carelessness. When the wrong answers were examined, it was observed that errors were gathered under the themes "lack of content knowledge" and "lack of reading comprehension" for verbal type input; under the theme "carelessness" for graphical type input; under the theme "lack of content knowledge" for algebraic and numerical type input.

The Application of Genetic Algorithm and an Evaluation Algorithm in Online Examination System

Online examination system plays a significant role in education. However, there are varieties of disadvantages in non-optimized systems, such as randomly selecting questions that make the exam paper has an imbalance difficulty, the unanticipated weight of knowledge points, and so on. A genetic algorithm is an efficient and achievable way to improve the ability to generate exam paper. Besides, a massive amount of data are generated when the system is running. Nevertheless, some of the systems only store the data, in another word, they do not make full use of the generated data. An evaluation algorithm is put forward in this essay to give objective and scientific evaluations on students’ learning and teachers’ teaching via using the data that are generated in examinations, which is based on the degree of difficulty. To make this algorithm working well, the degree of difficulty of questions stored in the database is supposed to be updated dynamically when the samples of questions’ answers become large enough.

Using Wearable Physiological Sensors for Affect-Aware Intelligent Tutoring Systems

Intelligent Tutoring Systems (ITS) have shown great potential in enhancing the learning process by being able to adapt to the learner's knowledge level, abilities, and difficulties. An aspect that can affect the learning process but is not taken into consideration by traditional ITS is the affective state of the learner. In this work, we propose the use of physiological signals and machine learning for the task of detecting a learner's affective state during test taking. To this end, wearable physiological sensors were used to record electroencephalography (EEG), electrocardiography (ECG), and electromyography (EMG) signals from 27 individuals while participating in a computerised English language test. Features extracted from the acquired signals were used in order to train machine learning models for the prediction of the self-reported difficulty level of the test's questions, as well as for the prediction of whether the questions would be answered correctly. Supervised classification experiments showed that there is a relation between the acquired signals and the examined tasks, reaching a classification F1-score of 74.21% for the prediction of the self-reported question difficulty level, and a classification F1-score of 59.14% for predicting whether a question was answered correctly. The acquired results demonstrate the potential of the examined approach for enhancing ITS with information relating to the affective state of the learners.

OPINI PESERTA SELEKSI CALON APARATUR SIPIL NEGARA (ASN) TERHADAP SISTEM COMPUTER ASSISTED TEST (CAT)

The purposes of using Computerized Assisted Test (CAT) system in civil servant selection were to create transparency, objectivity, accountability and efficiency. Some of the problems formulated in this study are how the public opinion of facilities and services; transparency and purity; and the level of difficulty of the CAT system exam questions. This study aims to determine the opinion of civil servant candidates for Bandung District about facilities, transparency and difficulty level of exam questions with the CAT system. The study was conducted in the Gedung Serbaguna Telkom University Bandung Regency. The population is all civil servant candidates who take the exam by CAT system in Bandung Regency in 2018. The number of samples is 100 people. The variables studied were facilities and services; transparency and purity; exam and graduation questions. Descriptive statistical analysis was performed with the help of SPSS 24 software. The results of the study obtained an average opinion of the participants on the facilities and service variables as well as transparency and were above 3 (three). The lowest average participant opinion is an indicator of the difficulty level of the questions (average = 1.91).

Pengembangan Asesmen Literasi Kimia pada Materi Hukum-Hukum Dasar Kimia dan Stoikiometri Kelas X SMA/MA

This chemistry literacy assessment instrument developed because of recent challenge toward student learning achievement evaluation so it could not only grade student’s cognitive ability in term of understanding and memorizing, but also will grade the application of students’ concept to face any problems. The availability of the chemistry literacy assessment instrument consists of various aspects, such as content, context, High Order Learning Skills (HOLS), and attitude. These aspects are expected to answer those challenges. This study aims to produce a proper and right chemistry literacy assessment toward fundamental law of chemistry and stoichiometry topic based on content validity value and questions numbers, reliability grade, difficulty level and chemistry literacy question differentiator level. This study is the development type, which is Model of Educational Reconstruction (MER) design. MER consists of three stages, (1) analysis of content structure, (2) empirical study, (3) development of instruction evaluation. The clarification of a test instrument on empirical study was done toward three chemistry and education experts (lecturer and teacher). The instrument used was chemistry literacy question which consists of seven discourse question, then expanded into 15 questions. The result showed that content validity value of the designed chemistry literacy assessment could be categorized as valid with 1,11 mark. From 15 questions, there are three questions in “very significant” category, nine in “significant” category and three other in “insignificant” category. The reliability test is 0,88. In difficulty level, there are eight questions in “medium” category, six in “difficult” and one in “very difficult”. The differentiator level, there are nine questions in “medium” category, four in “proper” and two other in “not proper”. Question in “insignificant” and “not proper” category is rejected. In conclusion, from 15 designed questions, there are 12 questions in “proper” and “right” category.

Pengaruh Model Pembelajaran Discovery Learning Terhadap Hasil Belajar Siswa Pada Mata Pelajaran Kearsipan di SMKN 10 Surabaya

This study aims to determine the effect of the discovery learning model of student learning outcomes in archival subjects in SMKN 10 Surabaya. This research is a quasi-experimental research design experiments with a form of nonequivalent control group design. This research was conducted at SMKN 10 Surabaya totaling 144 students. The sample in this study used students of class X OTKP 1 and X OTKP 2, both of whom had the same number of students as many as 36 students. The sampling technique used is non-probability sampling with purposive sampling types. The method in collecting data uses test, interviews, and documentation. This instruments used in this study used a syllabus, lesson plans, and pretest posttest sheets. Data analysis techniques in this study consisted of 2 parts including item analysis and analysis of student learning outcomes. Analysis of items including validity test, reliability test, the level of difficulty of questions, and dsitinguishing features. Where as in the analysis of student learning outcomes using homogeneity test, normality test, hypothesis testing, and gain score analysis. The results of this study prove that student learning outcomes in the cognitive domain increased by 86% in the experimental group compared to the control class with percentage 11%. Homogeneity test result produce P-Value (0,0854) &gt; significance level (0,05), which means that a sample can be said to be homogeneous. When viewed from the t-test posttest shows ttable with a confidence level of 95% (âˆ=5%) and df=70 then produces tcount (9,961) &gt; ttable (1,994), which indicates Ha accepted. When viewed from the t-test results showed the difference between pretest and posttest that tcount (15,838) &gt; ttable (1,994), which indicates Ha accepted. So it can be concluded that there is an influence of discovery learning model on student learning outcomes in subjects archieves at SMKN 10 Surabaya.

Efektivitas Analisis Butir Menggunakan Anajohn, Anates Dan Iteman Studi Soal Usbn Pelajaran Sejarah Kota Padang

This research aims to determine the quantitative analysis of the National Standard School Examination (USBN) questions on the history of Padang City High School in 2018/2019 school year using the Anajohn, Anates and Iteman programs. And to know the comparison of the effectiveness of programs in analyzing the quality of items and beneficial for the value of students. This research is quantitative descriptive. The subjects this research were students of class XII of SMA Negeri 2 Padang, SMA Negeri 4 Padang, and SMA Negeri 12 Padang. The data obtained were analyzed using three programs, namely Anajohn, Anates and Iteman. The results of the study showed: (1) Validity, in the Anajohn package A program there were 20 questions or 50% were declared valid and 18 questions or 45% were declared invalid, while there were 2 questions or 5% were not validated. In package B there are 30 questions or 75% declared valid and 10 items or 25% declared invalid. In the Anates package A program there are 30 questions or 75% declared valid and 10 questions or 25% declared invalid. In package B there are 27 questions or 67.5% declared valid and 13 items or 32.5% declared invalid; (2) Questions have high reliability; (3) The level of difficulty of easy questions; (4) Having good differentiation; (5) The function of the deception about the function; (6) An effective program for teachers to use is Iteman. While the benefit for the results or the value of students is Anates.

The Mind in the Machine: Mind Perception Modulates Gaze Aversion During Child–Robot Interaction

This study examined whether interacting with a humanoid robot influences children’s gaze aversion, an effortless strategy that people commonly use to facilitate thinking when asked challenging questions. Following the intentional stance model, we hypothesized that interacting with agents perceived as having a mind would modulate the social relevance assigned by the children to their interlocutor. Accordingly, we expected to observe an increase in children’s gaze aversion rates when questioned by an interaction partner believed to have a mind, compared to interaction conditions in which the questioner was believed to be a machine. To test this hypothesis, we involved 94 children in two experiments. In Experiment 1, the children interacted either with a humanoid robot (Human–Robot; n = 22) or with a human (Human–Human; n = 22) questioner. In Experiment 2, all the children interacted with a humanoid robot: one group was told the robot was controlled by a human (Avatar; n = 25), while the other group was told the robot was controlled by a computer algorithm (Machine; n = 25). Our results show that: (1) adopting an intentional stance (Human–Human; Avatar) increases gaze aversion rates; (2) gaze aversion increases and (3) response accuracy decreases as a function of question difficulty; (4) accuracy does not differ between interaction conditions. Based on these findings, we propose that gaze aversion rates might be considered an objective behavioural indicator of mind perception. Implications for robot-mediated education are also discussed.

Maximizing Computer Based Test (CBT) Teslet Model for Education Quality Mapping

Mapping the quality of education is one step that can be done to anticipate and overcome problems in the world of education. Basically, the mapping of the quality of education is aimed at equitable distribution of education. To achieve these objectives, it is necessary to maximize the development of three aspects of quality consisting of inputs, processes and outputs. The process as an important part of the responsibility of teachers has an extraordinary role in gathering various information or data needed to determine the quality or the quality of education at a particular institution or school. Therefore, one solution that can be taken which is to develop a test model based on computer, commonly known as Computer Based Testing (CBT). CBT is a form of software that can be developed using a computer. The development of the CBT in question must also be accompanied by a quality problem model so that the policies taken can be effective. The test model testlet is an alternative that can be maximized considering that this test is a collaboration between the form of problem and multiple-choice questions with the level of difficulty of questions that are tied. By using this testlet, the teacher can not only obtain information about student competencies, but can also obtain information about individual characteristics. In addition, the use of this test model will make it easier for governments and educators to conduct quality mapping because it is more effective and efficient

Dissociable mappings of tonic and phasic pupillary features onto cognitive processes involved in mental arithmetic.

Pupil size modulations have been used for decades as a window into the mind, and several pupillary features have been implicated in a variety of cognitive processes. Thus, a general challenge facing the field of pupillometry has been understanding which pupil features should be most relevant for explaining behavior in a given task domain. In the present study, a longitudinal design was employed where participants completed 8 biweekly sessions of a classic mental arithmetic task for the purposes of teasing apart the relationships between tonic/phasic pupil features (baseline, peak amplitude, peak latency) and two task-related cognitive processes including mental processing load (indexed by math question difficulty) and decision making (indexed by response times). We used multi-level modeling to account for individual variation while identifying pupil-to-behavior relationships at the single-trial and between-session levels. We show a dissociation between phasic and tonic features with peak amplitude and latency (but not baseline) driven by ongoing task-related processing, whereas baseline was driven by state-level effects that changed over a longer time period (i.e. weeks). Finally, we report a dissociation between peak amplitude and latency whereby amplitude reflected surprise and processing load, and latency reflected decision making times.

An Entropy Clustering Approach for Assessing Visual Question Difficulty

We propose a novel approach to identify the difficulty of visual questions for Visual Question Answering (VQA) without direct supervision or annotations to the difficulty. Prior works have considered the diversity of ground-truth answers of human annotators. In contrast, we analyze the difficulty of visual questions based on the behavior of multiple different VQA models. We propose to cluster the entropy values of the predicted answer distributions obtained by three different models: a baseline method that takes as input images and questions, and two variants that take as input images only and questions only. We use a simple k-means to cluster the visual questions of the VQA v2 validation set. Then we use state-of-the-art methods to determine the accuracy and the entropy of the answer distributions for each cluster. A benefit of the proposed method is that no annotation of the difficulty is required, because the accuracy of each cluster reflects the difficulty of visual questions that belong to it. Our approach can identify clusters of difficult visual questions that are not answered correctly by state-of-the-art methods. Detailed analysis on the VQA v2 dataset reveals that 1) all methods show poor performances on the most difficult cluster (about 10\% accuracy), 2) as the cluster difficulty increases, the answers predicted by the different methods begin to differ, and 3) the values of cluster entropy are highly correlated with the cluster accuracy. We show that our approach has the advantage of being able to assess the difficulty of visual questions without ground-truth (\ie, the test set of VQA v2) by assigning them to one of the clusters. We expect that this can stimulate the development of novel directions of research and new algorithms.

A comparative study of methods for a priori prediction of MCQ difficulty

Successful exams require a balance of easy, medium, and difficult questions. Question difficulty is generally either estimated by an expert or determined after an exam is taken. The latter provides no utility for the generation of new questions and the former is expensive both in terms of time and cost. Additionally, it is not known whether expert prediction is indeed a good proxy for estimating question difficulty. In this paper, we analyse and compare two ontology-based measures for difficulty prediction of multiple choice questions, as well as comparing each measure with expert prediction (by 15 experts) against the exam performance of 12 residents over a corpus of 231 medical case-based questions that are in multiple choice format. We find one ontology-based measure (relation strength indicativeness) to be of comparable performance (accuracy = 47%) to expert prediction (average accuracy = 49%).

How Effective are Matching Schemes in Enticing Low-income Earners to Save More for Retirement? Evidence from a National Scheme

Concerns over the adequacy of low and middle-income earner contributions to retirement plans have led governments to introduce targeted matching schemes. In this study, we examine the effects of a simple and generous Australian scheme using administrative tax-filer data, exploiting longitudinal changes in eligibility and match rates. We find small increases in the proportion who contribute and bunching at the eligible maximum, but lower average contributions because the matching payment displaces contributions of high contributors. Contributions through unmatched channels are also crowded out. These findings highlight the difficulties of targeting matching schemes and question the merits of simplifying them.