Svein Lie

Students’ Preference for Science Careers: International comparisons based on PISA 2006

This article deals with 15‐year‐old students’ tendencies to consider a future science‐related career. Two aspects have been the focus of our investigation. The first is based on the construct called ‘future science orientation’, an affective construct consisting of four Likert scale items that measure students’ consideration of being involved in future education and careers in science‐related areas. Due to the well‐known evidence for Likert scales providing culturally biased estimates, the aim has been to go beyond the comparison of simple country averages. In a series of regression and correlation analyses, we have investigated how well the variance of this construct in each of the participating countries can be accounted for by other Programme for International Student Assessment (PISA) student data. The second aspect is based on a question about students’ future jobs. By separating science‐related jobs into what we have called ‘soft’ and ‘hard’ science‐related types of jobs, we have calculated and compared country percentages within each category. In particular, gender differences are discussed, and interesting international patterns have been identified. The results in this article have been reported not only for individual countries, but also for groups of countries. These cluster analyses of countries are based on item‐by‐item patterns of (residual values of) national average values for the combination of cognitive and affective items. The emerging cluster structure of countries has turned out to contribute to the literature of similarities and differences between countries and the factors behind the country clustering both in science education and more generally.

PISA and scientific literacy: similarities and differences between the nordic countries

In this paper we have set out to search for similarities and differences between the Nordic countries concerning patterns of competencies defined as scientific literacy in the Programme for International Student Assessment (PISA) study. The first part focuses on gender differences concerning the two types of competencies, understanding of scientific concepts versus skills in scientific reasoning, based on analyses of sum scores of groups of items. The second part focuses on differences and similarities between countries based on item‐by‐item analyses. Correlations between each Nordic country (as well as the Nordic group as a whole) and every other country have been used to look for a Nordic pattern. In the last part cluster analysis has been used to see how countries establish clusters and whether these clusters represent meaningful groups in a geographical, cultural or political context.

Profiles of Students’ Interest in Science Issues around the World: Analysis of data from PISA 2006

The Programme for International Student Assessment in 2006 included several measures of students’ interest in science. These measures were constructed by combining information from several items where students are asked to respond to statements along Likert scale categories. Since there is evidence for Likert scales providing culturally biased country scores, we demonstrate in this article that the relative profiles of interest can be meaningfully analysed across countries. Hence, we have developed national relative profiles of interest in science constructed from the country‐ and item‐specific residuals at the item level. Subsequently, these relative profiles of interest have been used as input in a cluster analysis providing identification of distinct groups of countries with similar item‐by‐item patterns of interest in science. The most notable feature of the analysis is an overall division between two larger groups of countries, roughly corresponding to European/Western countries in one group and non‐European countries, with only a few exceptions, in the other group. A number of meaningful clusters of countries, partly defined by language and partly by localisation, are identified within each of the two main clusters. In order to develop a more detailed understanding of the characteristic features of the various clusters, descriptive information about the items is included in the analysis. The most notable finding is the strong relative preference for life and health issues among the non‐European countries, contrasted with the distinct favouring of items relating to physical/technological systems in the European/Western countries.

Learning about Students' Knowledge and Thinking in Science through Large-Scale Quantitative Studies.

The main issue addressed in this article is that there is much to learn about students’ knowledge and thinking in science from largescale international quantitative studies beyond overall score measures. Response patterns on individual or groups of items can give valuable diagnostic insight into students’ conceptual understanding, but there is also a danger of drawing conclusions that may be too simple and nonvalid. We discuss how responses to multiple-choice items could be interpreted, and we also show how responses on constructed-response items can be systematised and analysed. Finally, we study, empirically, interactions between item characteristics and student responses. It is demonstrated that even small changes in the item wording and/or the item format may have a substantial influence on the response pattern. Therefore, we argue that interpretations of results from these kinds of studies should be based on a thorough analysis of the actual items used. We further argue that diagnostic information should be an integrated part of the international research aims of such large-scale studies. Examples of items and student responses presented are taken from The Third International Mathematics and Science Study (TIMSS).RésuméL’article met à jour le fait que les études quantitatives internationales à grande échelle nous disent beaucoup, au-delà des mesures générales de scores, sur les connaissances et les pensées des élèves par rapport aux sciences. Les patterns des réponses données à des items individuels ou bien à des groupes d’items peuvent nous éclairer sur la compréhension conceptuelle des élèves, mais risquent de conduire à des conclusions trop simples et non-valides. On discute les moyens de systématiser et d’analyser les réponses à des items à choix multiple. Enfin, on étudie empiriquement les interactions entre certains caractéristiques des items et les réponses des élèves. On montre qu’il suffit d’une petite modification dans la manière de formuler et/ou formater l’item pour produire un effet substantiel dans le pattern de la réponse. Il s’ensuit que l’interprétation des résultats provenant de ce type d’études doit être basée sur une analyse approfondie des items administrés. Il s’ensuit également que l’information diagnostique peut être une partie constitutive des objectifs de recherche internationaux qu’on cherche à réaliser par ce type d’études à grande échelle. Les exemples présentés d’items et de réponses des élèves sont extraits de la Third International Mathematics and Science Study (TIMSS).

Nordic PISA 2000 in a Sociocultural Perspective.

The Programme for International Student Assessment (PISA) is an international study coordinated by governments of participating countries, through the Organisation for Economic Co-operation and Development (OECD). In 2000, a total of 265,000 students from 32 countries took part in PISA. The main aim of PISA is to assess how well 15-year-old students are prepared to meet the challenges of today’s and tomorrow’s knowledge societies. The assessment is forward looking, focusing on young people’s ability to use their knowledge and skills to meet real-life challenges, rather than mastering a specific school curriculum. Assessing young people at the end of their compulsory education is regarded as providing valuable insights into the performance of basic education systems (OECD, 1999, 2001). PISA 2000 covered three domains: reading literacy, mathematical literacy and scientific literacy. In 2000 priority was given to reading literacy, with mathematical and scientific literacy assessed in lesser depth. PISA will continue in three-year cycles, the focus shifting between the three domains. In 2003 the focus was on mathematical literacy, and in 2006 it will be on scientific literacy. In the future, with the collection of data in the same domain every three years, it will be possible to examine trends in student performance over time. In addition to assessing students’ performance in the three core domains, PISA examines competencies across disciplinary boundaries such as students’ motivation, attitudes and learning strategies, as well as comfort with and perceived ability to use computers. Furthermore, various background factors such as students’ socio-economic and cultural home background, family structure and immigration status, educational resources and cultural activities outside school are surveyed (OECD, 2001).

Timss Science Results Seen from A Nordic Perspective

In this chapter we describe some characteristic features of the TIMSS science results as seen from a Nordic perspective, and in particular; Norway. Four out of the five Nordic countries participated in TIMSS: Iceland and the three Scandinavian countries, Denmark, Norway, and Sweden. (Finland did not take part, but participated as the only Nordic country in the repeat of TIMSS in 1999.) These four Nordic countries have much in common historically, culturally and politically. In the Scandinavian countries almost the same language is spoken, whereas Icelandic is different, but quite similar to the old common Nordic language. During the last six or seven hundred years Sweden and Denmark have rivaled each other for hegemony in the area, while Norway, Iceland, and Finland have been the smaller brothers in union with one or the other for most of the time. Today all Nordic countries are independent democracies with a strong social democratic tradition. They are relatively homogeneous societies with highly developed welfare systems. There are also strong and friendly cultural links among the Nordic countries, even if the European Union has split these countries into members (Denmark, Finland, and Sweden) and non-members (Iceland and Norway). Sweden has the largest population, eight million, whereas Denmark, Finland, and Norway all have between four and five million. Iceland has a small population, about 200,000. It should also be mentioned that the climate is similar in all of these countries. Denmark stands out from the other Nordic countries in two respects: firstly, due to its position closer to the other North European countries it appears more “continental” both in climate and culture; and secondly, its relatively high population density makes Denmark somewhat different from the typical Nordic countries which have wilderness within easy reach of even the largest cities.

Country profiles of scientific competence in TIMSS 2003

The aim of the present contribution is to investigate similarities and differences of strengths in science competences between countries, based on TIMSS 2003 data. Analyses are based on systematic investigation of patterns of p values for individual science items. Hierarchical cluster analysis was applied to establish meaningful groups of countries. The resulting pattern of how countries cluster together into groups of increasing size is presented and discussed. The features for each group of countries are described in terms of relative strengths according to item format, subject domain, and cognitive domain. Finally, the measures on relative strengths in subject domain were compared to the relative emphases in the intended curriculum and in the implemented curriculum (percentage of topics taught in the classroom). It turned out that the data on emphases in the classroom could explain more of the relative strengths and weaknesses than the intended curriculum.

How Can Large International Comparative Studies Contribute to the Quality of Science Education

In this paper the two international comparative studies IEA TIMSS and OECD PISA have been discussed by comparing their similarities and differences. A number of examples have been presented to demonstrate how findings in various areas are relevant to help improve science education. Focus are on students’ conceptual understanding, gender and school differences, relations to home background factors, and on what characteristics of instruction that seem to be related to high achievement. Furthermore, the assessment frameworks for the two studies are argued to be of influential importance in its own terms, but that any influence on national aims and curricula should be carefully considered only in a national context.