This mixed methods study is based on TIMSS 1999 data, where two-level (student and school) HLM-models of students' science achievement have been created. The study has practical approach, which means that each country has its own model based only on statistical significant explanatory variables. Explanatory variables are presented in a new graphical form, which makes results to be easier interpreted by persons unfamiliar to HLM-modelling.First outcomes of this study are country specific HLM-models. These models have received 'emic'1) interpretations by science education experts from studied countries.Four questions were formulated to guide the discussion of the study1) How well does this multilevel modelling approach explain science achievement in the countries studied? In this part of the study, the development of percentage explained for within-school, between-school, and total variance is revealed step by step for each country.2) How well do the significant explanatory variables of science achievement identified in this study work in other countries and what is the relationship of these items to science achievement in other participating countries? 3) How applicable is the approach used in this study to reveal explanatory variables in different science subjects and domains of TIMSS?As results, this study gives some explanations why science, compared to mathematics, has been in minor role of further analysis of LINCAS data.References:Kupari, P. & Reinikainen, P. 2004 Matematiikan ja luonnontieteiden osaaminen Suomessa vuosituhannen vaihteessa: TIMSS 1999-tutkimus. [How Finnish students achieve in mathematics and science in turn of the century] In K. Leimu. Kansainväliset IEA-tutkimukset Suomi-kuvaa luomassa Koulutuksen tutkimuslaitos, Jyväskylän Yliopisto. 305-334. Kupari, P. Reinikainen, Nevanpää, T. & Törnroos, J. 2001. Miten Matematiikkaa ja luonnontieteitä osataan suomalaisessa peruskoulussa? [National report of TIMSS 1999 in Finland]. Institute for Educational Research. University of Jyväskylä. Park, C. & Park, D. 2006. Factors Affecting Korean Studentsä Achievement in TIMSS 1999. In S. J. Howie & T. Plomp (Eds.) Contet of Learning Mathematics and Science. Lessons learned from TIMSS. Routledge: Taylor & Francis Group. London, New York. 177-191.