Scoring Method of the TSA Thinking Assessment

NDO - A notable point in the scoring method of the TSA Thinking Assessment of Hanoi University of Science and Technology is that it does not use the traditional scoring method but applies the question response theory in calculating the candidate's score to be able to estimate the candidate's ability accurately and reliably.

The Thinking Skills Assessment (TSA) test of Hanoi University of Science and Technology aims to assess three basic thinking abilities of students, including: Mathematical Thinking, Reading Comprehension and Scientific Thinking/Problem Solving - important factors that help students to study best in a university education environment.

Accordingly, the test consists of 3 parts: Mathematical Thinking (60 minutes), Reading Comprehension (30 minutes) and Scientific Thinking/Problem Solving (60 minutes). These are 3 independent parts, the test questions will focus on assessing the thinking ability of the candidate in each part, not directly testing the knowledge of any subject.

The TSA exam consists of three independent sections (Mathematical Thinking, Reading Comprehension, Scientific Thinking/Problem Solving) that focus on assessing candidates' abilities, without directly testing knowledge of any subject.

There are a few points worth noting about the scoring method of the TSA Aptitude Test.

The Admissions and Career Guidance Department of Hanoi University of Science and Technology said that the traditional scoring method is to use raw scores to evaluate the final exam. According to classical multiple choice theory, the raw score of the exam is the sum of the scores of the questions in that exam.

For example, in a test with 100 questions, each correct answer will earn the candidate 1 point. If the candidate answers 70 questions correctly, he/she will receive 70 points. This score is called the raw score and is used in evaluating the results as well as comparing with the scores of other candidates.

One of the disadvantages of using raw scores is that it is difficult to distinguish the ability of candidates with the same raw score when taking the same test at the same time. In addition, when calculating ability scores using raw scores, the ability of candidates will change when taking two tests of different difficulty at the same time. For example, in a test session, candidates who answer 70 questions correctly will have the same raw score of 70 points.

In reality, these candidates may have answered different sets of questions correctly and the difficulty of these questions in the exam also varies so a score of 70 does not accurately reflect the ability of these candidates.

To address this issue, major exams around the world have adopted modern educational measurement theories to more accurately and reliably estimate candidates' abilities. One of them is the application of question response theory. This theory hypothesizes that each candidate answers a question in a test with a certain level of ability and that candidates with high ability will have a higher probability of correctly answering any question than candidates with low ability. Based on this theory, parameters such as the difficulty and discrimination of questions as well as candidates' abilities can be quantified and these factors are independent of each other.

This means that the parameters are specific to the question, independent of the sample of candidates answering that question, and conversely, the candidate's ability is invariant to the questions they answer. Therefore, the estimate of candidate ability will be more reliable than the raw score method.

Based on the test results of the candidates, the scoring algorithm will determine the thinking level of each question according to the percentage of candidates who answer the question correctly. According to the principle, questions with few correct answers will be questions with a high level of thinking, questions with many correct answers will be questions with a low level of thinking.

The TSA Thinking Assessment Test of Hanoi University of Science and Technology uses the question response theory in calculating the TSA scores of students. With this calculation, although candidates have the same raw score, based on the difficulty level of the questions that each candidate can answer, the corresponding ability level will be estimated, then this score will be converted to a 100-point scale.

The scoring scale will be formed after the test results of the candidates after each test. Based on the test results of the candidates, the scoring algorithm will determine the level of thinking of each question according to the percentage of candidates who answer that question correctly. According to the principle, questions with few correct answers will be questions with a high level of thinking, questions with many correct answers will be questions with a low level of thinking. The scoring scale is formed according to the principle that questions with a high level of thinking will be scored high in proportion to the percentage of candidates who answer correctly, questions with a low level of thinking will be scored low in proportion to the percentage of candidates who answer correctly.

In addition, to ensure fairness between exam sessions, the TSA exam of Hanoi University of Science and Technology also applies a set of standard "bridge" questions - also known as bridge questions, common questions - between exam papers to bring the exam assessment indicators to the same scale. This technology will ensure that candidates in different exam sessions are assessed on the same common scale, thereby ensuring fairness in results between exam sessions.

During the 2024 university admission season, nearly 50,000 times, with a total of about 21,000 candidates, took the TSA Thinking Assessment test of Hanoi University of Science and Technology. The test results were used by more than 50 universities to serve the regular university admission work. In 2025, the Thinking Assessment tests are expected to be held from January to April.

Source: https://nhandan.vn/phuong-phap-cham-diem-cua-bai-thi-danh-gia-nang-luc-tsa-post845455.html