The rapid changes in the times we live have led to an increase in the importance of scientific skills in our lives. To overcome the challenges of the twenty-first century in the science and technology sector, students need to be equipped with 21st-century skills to ensure their competitiveness in the globalization era (Turiman et al., 2013).

Among the 21st-century skills, the most important ones are numeracy and scientific literacy (Word Economic Forum, 2015; OECD, 2013). Scientific literacy is the ability to engage with science-related issues, and with the ideas of science, as a reflective citizen (OECD, 2017). It emphasizes the importance of being able to apply scientific knowledge in the context of real-life situations. Numeracy represents the ability to use numbers and other symbols to understand and express quantitative relationships (World Economic Forum, 2015).

TIMSS is the most advanced study that can provide an overview of the results of Romanian eighth-grade students in mathematics and sciences (physics, chemistry, biology, and geography). In 2019, Romanian students obtained a score of 479 points in mathematics (intermediate international benchmark) and 470 points in science (intermediate international benchmark). Analyzing Romania's participation (2007 - 2019) in the TIMSS study, it can be seen that the mathematics scores situate within the range of 458-479, and the science scores situate within the range of 462 - 470.

Unfortunately, the results obtained by the Romanian students within TIMSS 2019 remain below the average of the European countries and far below the regional average, being observed a significant variation in the quality of the national education system: the percentage of students who obtained "advanced" results is only 6% in mathematics and 4% in physics, while the percentage of students with "low" results or below the average-functional level is 70%. Romania recorded a much higher rate than other countries in terms of numerical or scientific illiteracy: 22% of students were not able to use mathematics or physics even in the simplest contexts.

 The proposed research investigates the factors that affect the learning process in mathematics and sciences for 8th-grade students. Among the learning factors we will take into consideration, we mention: (1) carrying out experiments during science classes, (2) the way of working in the classroom (teamwork, individual work), (3) frequency of homework, (4) allocated time for homework, (5) self-efficacy towards math and science, (6) positive affect towards math and sciences, (7) teaching methods, (8) private lessons and also demographic characteristics as (9) gender and (10) residence.

 The data analysis procedure will be conducted in two steps: (1) Analysis of each predictor’s (1-10) contribution to the total variance of TIMSS results of participants; (2) Comparison between high and advanced benchmark students (highest 25% of scores) and low and intermediated benchmark students (lowest 25% of scores) taking into consideration all the predictors, to see which of them contributes most to the results of high and advanced benchmark students.

The study sample was established following a random probability sampling process. All the schools in Romania that had the eighth grade in their composition were taken into consideration, each school having an equal chance of being chosen. There had been used also the following exclusion criteria: (1) schools operating according to a different curriculum (15 schools), (2) schools with special needs children (243 schools), (3) very small schools (449 schools). To increase the representativity of the sample, two layers were used in the selection of schools: (1) the environment of origin with two categories: rural and urban, and (2) the geographical region with five regions. Following this sampling process, a sample consisting of 199 public schools resulted. From these schools, 4,485 students (14-15 years) participated in the study. Most of the schools participating in the study are located in small towns or villages (40.7%), followed by those in the urban area (26.3%), the suburban area (9.8%), respectively the rural area, with difficult access (7.2%). Data collection was carried out through two methods: administering tests to students in mathematics and sciences and the administration of context questionnaires to students. All test booklets and context questionnaires were applied on the same day. Firstly, the test booklets were applied and then the context questionnaires. During the test period, the students were supervised by a teacher who didn’t have classes with the tested students. The tests administered to students included multiple-choice items and constructed responses. The test items were distributed in 14 test workbooks and each test workbook included 28 math items and 28 science items. Context questions provide information that helps interpret the results of math and science tests. The students answered questions related to the teaching methods used by teachers in the classroom, the way mathematics and science lessons are conducted, as well as factors related to the preferences for mathematics and science or the positive affect. The data analysis is based on a statistical approach and between the methods proposed to be used we mention multiple regression (to analyze the contribution of the predictor variables to the total variance of TIMSS results), hierarchical multiple regression (to see in which measure the learning factors predict the TIMSS results under controlling for the influence of other factors) and relative predictor weight (to calculate the relative importance of predictor variables in contributing to TIMSS results).

Following some preliminary analyses, we noticed that self-confidence in classes is a variable with a strong effect that predicts results in mathematics. We also observed that the duration and number of private lessons, positive affect towards mathematics, and duration of homework are medium-effect predictors for mathematics achievements. Individual work in mathematics classes is a variable with a negative effect, the higher the value, the more negatively it affects school performance in mathematics. In sciences, we observed that carrying out experiments during science classes has a strong effect that predicts results in sciences and individual work in science classes is a variable with a negative effect, the higher the value, the more it negatively influences school performance in science. TIMSS 2019 results offer a strong basis for decision-making based on scientific evidence to improve educational policies and practices related to teaching and learning mathematics and sciences. Based on the national results, they can be identified the leading teaching and learning styles addressed in mathematics and sciences can be captured with objectivity the less effective learning methods and cognitive strategies used by students. Situational factors can be identified that have an impact on learning performance. This information can and should be of great importance for educational policies that promote equity and equal opportunities in education. Through this research, we hope to come to the aid of teachers with results that will help them to make their teaching methods more efficient in the classroom in order to improve the results of students in mathematics and science, thus making it possible to increase the advanced benchmark of students in Romania.

Ciolan, L., Iliescu, D., Iucu, R., Nedelcu, A. Gunnesch-Luca, G. (coord.) (2021). Romania in TIMSS: Country report. https://unibuc.ro/wp-content/uploads/2021/06/TIMSS-Raport-de-tara-2021-05-07.pdf Griffin, P., & Care, E. (2015). Assessment and teaching of 21st century skills: Methods and approach. Springer. Maass, K., Geiger, V., Ariza, M.R. & Goos, M. (2019). The Role of Mathematics in interdisciplinary STEM education. ZDM Mathematics Education 51, 869–884. https://doi-org.am.e-nformation.ro/10.1007/s11858-019-01100-5 Organization for Economic Co-operation and Development (OECD). (2013). OECD skills outlook 2013: first results from the survey of adult skills. Paris: OECD Publishing. Organization for Economic Co-operation and Development (OECD). (2019). Future of Education and Skills 2030. Concept Notes. https://www.oecd.org/education/2030- project/contact/OECD_Learning_Compass_2030_Concept_Note_Series.pdf Partnership for 21st Century Learning. (2015). P21 framework definitions. Retrieved from http://www.p21.org/documents/P21_Framework_Definitions.pdf. TIMSS. (2019). Encyclopedia: Education Policy and Curriculum in Mathematics and Science, Romania. https://timssandpirls.bc.edu/timss2019/encyclopedia/romania.html TIMSS. (2019). Assessment Frameworks. https://timssandpirls.bc.edu/timss2019/frameworks/ Turiman, P., Omar, J., Daud, A. & Osman, K. (2012). Fostering the 21st Century Skills through Scientific Literacy and Science Process Skills. Procedia - Social and Behavioral Sciences. 59. 110–116. https://www.sciencedirect.com/science/article/pii/S1877042812036944 World Economic Forum. (2015). New vision for education: unlocking the potential of technology. Geneva: World Economic Forum.