A Scientific Analysis of the Reliability of Multilevel Assessment of Speaking Skills

The article examines the reliability of the Multilevel Speaking Test administrated by the Knowledge and Skills Assessment Agency under the Ministry of Higher Education, Science and Innovation of the Republic of Uzbekistan. Assessing speaking test – oral proficiency is essential in language testing, but it can be challenging due to various factors influencing evaluations. The piece delves into reliability in language testing, emphasizing the importance of consistent assessment outcomes. It explores the role of raters, discussing intra-rater and inter-rater reliability, and how subjective evaluation can impact test scores.
The article is about the Multilevel Speaking Test, an English proficiency examination aiming to evaluate students’ language levels in Uzbekistan’s educational system. The test is a criterion-referenced proficiency examination and integral to the Common European Framework of Reference for Languages (CEFR). The speaking section of the test is discussed in detail, including its structure, evaluation framework, and factors that affect its reliability. Rater variability and test conditions are two critical aspects that are evaluated to ensure the test’s reliability. The test employs two raters to mitigate subjectivity, but deficiencies in rater training and cultural biases may affect reliability. The article also highlights the benefits of computer-delivered tests for reliability, but it also points out issues such as limited sample materials and the singular task format that may impact the test-taker’s performance. In conclusion, the Multilevel Speaking Test is reliable because it uses two raters, but the article suggests that rater training should be improved to reduce biases. The reliability of the test can be further improved by providing more sample materials, adding more tasks, and improving test conditions.
After a comprehensive evaluation, the Multilevel Speaking Test was found to be reliable. However, there is room for improvement in rater training, test conditions, and sample availability to enhance its effectiveness in measuring oral proficiency.