You are here
Articles
number: /2024
CIT number: 0
Number of views: 1,099
Teachers’ knowledge and skills in assessment are important factors for improving the practice of Classroom Assessment (CA), which is a vital component of teaching and learning. There is ample research attempting to ascertain if training can contribute to increasing teachers’ knowledge and skills in assessment. In Bangladesh, the new curriculum places significant emphasis on classroom assessment compared to summative assessment. This study aims to answer the following questions: What is the status of Classroom Assessment Literacy (CAL) among primary school teachers in Bangladesh? What is the relationship between teachers’ training in assessment and their CAL? And what factors do teachers believe affect the translation of training in CA into CAL? The study followed a mixed-method research approach. Initially, quantitative data on teachers’ CAL and their training in CA were collected from 100 primary school teachers in Dhaka city using a CAL Test. Based on the results of this test, eight teachers with different combinations of CAL levels and training were interviewed. The results showed that the mean score of the sample of primary school teachers was above the middle score but still not high. Their mean score varied across different components of the test. For example, teachers were found to be most competent in using assessment data to make decisions and least competent in selecting appropriate assessment tools. We found that the more hours of training in assessment the teachers had, the lower they scored on the CAL test. However, this relationship was weak. We also found that having a Master of Education (MEd) degree can make a significant difference in their CAL test score. Teachers with more years of experience appeared to score lower on the test and vice versa (though not statistically significant). Qualitative data revealed that each teacher followed their own unique CA styles with a combination of different CA techniques, and there was no pattern based on teachers’ CAL test scores. However, teachers with higher CAL test scores seemed to provide feedback to their students in a more descriptive and meaningful way than teachers with lower CAL test scores. Data revealed that training is one of the sources of teachers’ knowledge and skills in CA among others. However, various enabling and deterring factors work together to determine the effectiveness of training in bringing about sustainable change in teachers’ knowledge and skills in CA. Among many factors, the “one shot” and “one size fits all” design of the trainings were found to be ineffective by teachers who have unique classroom contexts. Additionally, teachers’ own perceptions of training, the school environment, and context are also factors that affect the translation of training in CA into sustainable CAL that they can apply in practice.
number: /2023
CIT number: 0
Number of views: 775
Prior research has demonstrated that different noncognitive factors influence students’ science literacy. The present research investigates the relationships among students’ perceptions of their science teachers, their non-cognitive outcomes (epistemological beliefs, science interests and usefulness of science) and the influence of these factors on students’ science literacy. The data includes 5,826 15-year-old students (52.2% male and 47.8% female) who participated in the Programme for International Student Assessment (PISA) in 2015. The research revealed a meaningful pattern of complex relationships among non-cognitive factors and their influence on students’ science literacy, enhancing and clarifying previous research findings with both theoretical and practical significance. The results of confirmatory factor analysis show that all items in each non-cognitive outcome had reasonable factor loading, and the model had good fit indices [RMSEA = 0.037; CFI = 0.931: TLI = 0.923; SRMR = 0.032]. The results of structural equation modeling (SEM) show good fits, suggesting that students’ science interests directly influence their science literacy. Furthermore, students’ perceptions of their science teachers, epistemological beliefs and usefulness of science indirectly influence science literacy through their science interests. Evidence from PISA Vietnam 2015 data also shows the insignificant path of the usefulness of science to science literacy.
number: /2023
CIT number: 0
Number of views: 939
A growing number of studies utilising the Rasch model have been published, spanning various disciplines. This paper delves into five key domains where Rasch measurement is applied, namely: developing new instruments, creating short-form instruments, developing vertical scales, combining Rasch analysis and path analysis, and applications to classroom testing. Within each area, this paper presents research studies as illustrations of how Rasch measurement can effectively address measurement issues and advance practices. These studies provide concrete examples for reflection, highlighting typical procedures and potential pitfalls to avoid when employing the Rasch model for different purposes in diverse contexts.
number: /2023
CIT number: 0
Number of views: 952
This study aimed to calibrate items aiming at measuring English reading comprehension ability in students who learn English as a second language. A total of 571 multiple-choice items were administered to 466 participants from 14 schools via a computer-adaptive test. These participants were mainly secondary school students. Data were analyzed using the Rasch model for dichotomous items. Results indicated that the instrument was sufficiently unidimensional and was quite well targeted at the students. It was able to measure the English abilities of secondary school L2 students. Item measures were also compared to the expert’s levelling of item difficulty levels. Based on the Pearson Correlation coefficient of 0.77, the items demonstrate a moderate shared variance, indicating a reasonably positive correlation with the expert’s levelling.
number: /2023
CIT number: 0
Number of views: 805
This paper contrasts accountability focussed Rasch based assessment systems with those that are aimed at supporting student learning, by drawing at distinctions between assessment of learning and assessment for learning. It includes a modest proposal for implementing a Rasch based assessment system in developing countries. After highlighting problems with current educational testing regimes, it describes two exemplary assessment systems which provide good models for developing testing in developing countries. In particular, it canvasses specific recommendations which move away from testing of learning to testing for learning. "If I had to reduce all of educational psychology to just one principle, I would say this. The most important single factor affecting learning is what the learner already knows. Ascertain this and teach him accordingly" (Ausubel, 1968).
number: /2023
CIT number: 0
Number of views: 771
It is important for teachers to be experts in the content that they teach, and it will be helpful for professional development programmes to design suitable and fair assessments for teachers. The purpose of this study was to use Rasch analysis to improve the functioning of the items in an instrument designed to assess the trigonometry knowledge of mathematics teachers. The participants in the study were 168 high school mathematics teachers who were enrolled in an in-service programme. Fourteen items were rescored to resolve the disordered thresholds that they exhibited. The rescoring resulted in an improved fit of the instrument. Furthermore, this cohort of teachers struggled with higher-level trigonometry questions, especially those which required shifts between different registers of representations. It suggests that the Department of Education needs to provide support to teachers by offering workshops which focus on building up their knowledge in trigonometry by also developing their representational fluency in trigonometry.
number: /2023
CIT number: 0
Number of views: 754
Our 18th PROMS meeting took place in Hanoi near the end of 2022. Now, I am delighted to report that, less than 12 months later, we are able to disseminate several key papers from that meeting to a wider audience in this special PROMS edition of the Vietnam Journal of Educational Sciences. I initiated the Pacific Rim Objective Measurement Symposium in 2005 with the express purpose of mentoring the growth of Rasch measurement in developing countries on the western Pacific rim. The focus has been to encourage the adoption of equitable scientific measurement into university communities and to those responsible for educational and other testing in the region.
number: /2023
CIT number: 0
Number of views: 2,642
Vietnam’s 2018 general education curriculum has been developed according to the competency approach. As a result, the instruction and assessment systems need to be adapted to align with the requirements of the new curriculum. Experience from other countries shows that one effective solution is applying psychometric theories in different assessment programs. In this paper, the educational assessment system in Vietnam will be introduced along with the history of psychometric theories, especially Rasch measurement, in Vietnam. After rehearsing some illustrated studies using psychometric theories as well as identifying some challenges of this field in the context of curriculum innovation, a proposal for developing instruments using psychometric theories will be discussed in relation to the requirements of assessment in the new general education curriculum in Vietnam.
number: /2023
CIT number: 0
Number of views: 911
Conducting assessment is well-known for posing serious challenges to teachers, particularly with the pluralism in assessment approaches. In Vietnam, English language teachers recently are required to master specified pedagogical competencies, including assessment of students’ newly defined language competence, and the national English curriculum has been continuously updated, while most teachers are not prepared for conducting classroom assessment in a competence-based approach. Local English teacher training universities have responded to this need by implementing language assessment courses. The conceptualization of pre-service teachers’ assessment competence and the effects of these courses, however, remains fairly unknown. This study examines the construct of preservice teachers’ assessment competence and the results of one introductory language assessment course at a leading language teacher education institution in Vietnam, focusing on measuring the competence of teachers-to-be in language assessment. Both Rasch analysis and descriptive statistics were used to analyze the functioning of the items in a localized construct of assessment literacy for preservice English teachers and these teachers’ self-assessment results, which revealed that a) many practical components of the construct of language assessment competence can function well, but some requirements need to be adapted and instructed with more attention for students in the future courses, and b) while many student teachers could sufficiently perform the required language assessment tasks and were fundamentally ready for the job, others faced serious difficulties in applying the more abstract principles. The findings shed light on the necessary adjustments to be made in defining assessment literacy for the target teachers and offer suggestions for the future professional development of both these teachers and in-service ones.
number: /2023
CIT number: 0
Number of views: 784
In many studies in Malaysia, several issues in providing evidence toward construct validity have been observed. In most problem cases, studies only provided face validity, factor analysis and reliability index yet claimed their instruments have sufficient construct validity. Another issue was a mix-up on assessment-type and perception-type items. Finally, insufficient sampling and targeting during pilot fail to provide empirical evidence on content validity. This paper presents the construct validity requirements according to Messick’s construct validity framework and proposes several methods to deal with the above issues.